Lateral Load Test of Drilled Shaft in Clay Reveals Structural Failure

TECHNICAL REPORT STANDARD TITLE PAGE
1. Report No. 2. Government Aceeuion No. 3. Recipient' 1 Catalog No.

FHWATX77 -211 -1
4
~;~t~eR;~s~~i~~ ~EST OF A DRILLED SH~-FT-.-_I_N_C-LA_Y__,______-+:::---:NR.=-~~-oe-,~-=~-o~-~-.-1-9_7_7 --_-._-_-__-------~1
6. Performing Orgoniaotoon Code
7. Autlt6t1 i) 8., ?erformong Orgoni aohon Report No.
Vernon R. Kasch, Harry M. Coyle, Richard E. Research Report 211-1
Ba rtoskewi tz, and Wi 11 iam G. Sa rver
9, Performing Organization Nome and Address 10. Work Unit No.
Texas Transportation Institute
~1~1-.-=C-on-tr-oe-to-r-:G-,o-nt~N~o.----------~
Texas A&M University
College Station, Texas 77843
~12-.-S-po-n-so-rin_g_A-ge-n-cy_N_a-me_o_n_d-Ad-d-re_ss_______________________~~
13. Type of ReportSeptember,
ond Period Covered
1976 II
Texas State Department of Highways and Public Trans- Interim -November, 1977 :
portation; Transportation Planning Division ~-:--:-~----,-----=--:----------<'
p, 0. BOX 5051 14. Spontoring Agency Code 1
Austin, Texas _J87~3~----------------~-------~-------~----------l!

15. Supplementary Notes 1
Research performed in cooperation with DOT, FHWA. \
Research Study Title: Design of Drilled Shafts to Support Precast
Panel Retaining Walls.
16. Abstract I
The behavior of a laterally drilled shaft in clay has been investigated by !
conducting a lateral load test on an instrumented shaft. Lateral deflection, shaftl
rotation, and soil resistance were measured for each applied load. Dial gages
were used to measure lateral deflection, while the shaft rotation was determined
by means of an inclinometer. Pneumatic pressure cells were installed in the shaft
at various depths to measure the soil resistance. The applied lateral load was
measured with a strain gage load cell.
Structural failure of the shaft occurred before the soil failed and prevented
the attainment of the ultimate lateral soil resistance. However, it was possible.
to compare the ultimate soil reactions predicted by several analytical procedures
with the soil reaction calculated from the maximum recorded soil resistance. Also,
an ultimate lateral load for the test shaft was predicted by various analytical ,
methods. A comparison was made between the maximum recorded load and the various ,.
predicted ultimates. Finally, a comparison was made between two ultimate test -1
loads obtained from the technical 1 iterature and the ultimate loads predicted by
the analytical methods.
I
A tentative procedure for the design of the rigid, laterally loaded drilled 1
shafts is presented. However, additional load tests are needed before the design j
can be finalized.
I
I
I
17. KeyWords 18. Distribution Statement

i
!I
Drilled Shaft, Lateral Load Test, Pres- No restrictions. This document is i
\ sure Cells, Soil Reaction, Ultimate Load available to the public through the i
Deflection, Rotation, Soil Creep. National Technical Information Service, I
Springfield, Virginia 22151
19. Security Clauil. (of this-report) Clouil. (of thit page) 21. No. of Pages Price
~
urity 22.
Unclassified classified . 97 /
L----------------------
Form DOT F 1700.7 1&&91
---------------------L--------~-----------J
LATERAL LOAD TEST OF A DRILLED SHAFT IN CLAY
by
Vernon R. Kasch
Research Assistant
Harry M. Coyle
Research Engineer
Richard E. Bartoskewitz
Engineering Research Associate
and
William G. Sarver
Research Associate
Research Report No. 211-1
D~sign of Drilled Sh~fts to Support Precast Panel Retaining Walls

Research Study Number 2-5-77-211
Sponsored by
State Department of Highways and Public Transportation
in Coop~ration with the
U.S. Department of Transportation
Federal Highway Administration
November 1977
Texas Transportation Institute

Texas A&~1 University
Co)lege Station, Texas
Disclaimer
The contents of this report reflect the views of the authors

who are responsible for the facts and accuracy of the data pre-
sented herein. The contents do not necessarily reflect the views
or policies of the Federal Highway Administration. This report
does not constitute a standard, specification or regulation.
There was no invention or discovery conceived or first
actually reduced to practice in the course of or under this con-
tract, including any art, method, process, machine, manufacture,
design or composition of matter, or any new and useful improve-
me.nt thereof, or any variety of plant which is or may be patent-
able under the patent laws of the United States of America or any
foreign country.
ii
ABSTRACT
The behavior of a laterally loaded drilled shaft in clay has been

investigated by conducting a lateral load test on an instrumented shaft.
Lateral deflection, shaft rotation, and soil resistance were measured
for each applied load. Dial gages were used to measure lateral deflec-
tion, while the shaft rotation was determined by means of an inclinometer.
Pneumatic pressure cells were installed in the shaft at various depths
to measure the soil resistance. The applied lateral load was measured
with a strain gage load cell.
Structural failure of the shaft occurred before the soil failed and
prevented the attainment of the ultimate lateral soil resistance. How-
ever, it was possible to compare the ultimate soil reactions predicted
by several analytical procedures with the soil reaction calculated from
the maximum recorded soil resistance. Also, an ultimate lateral load
for the test shaft was predicted by various analytical methods. A
comparison was made between the maximum recorded load and the
various predicted ultimates. Finally, a comparison was made betweentwo
ultimate test loads obtained from the technical literature and the
ultimate load's predicted by the analytical methods.
A tentative procedure for the design of rigid, laterally loaded
drilled shafts is presented. However, additional load tests are needed
before the design can be finalized.
KEY WORDS: Drilled Shaft, Lateral Load Test, Pressure Cells, Soil
Reaction, Ultimate Load, Deflection, Rotation, Soil Creep.
iii
IMPLEMENTATION STATEMENT
The results of the lateral load test and the state of the art
survey of the technical literature were combined to produce a tenta-
tive procedure for the design of drilled shafts supporting precast
panel retaining walls. An important feature in the design of retaining
walls is the allowable amount of wall deflection. Consequently, factors
of safety were included in the tentative design procedure to control
the amount of wall deflection by limiting the shaft rotation.
The tentative design procedure could be implemented but additional
testing would undoubtedly result in improvements in the procedure.
Future lateral load tests should be conducted to failure in order to
obtain ultimate loads and ultimate soil reactions. A maximum value
for drilled shaft rotation also needs to be established, and a study
of soil behavior under the influence of sustained lateral loads needs
to be performed. Finally, load tests on shafts of various depths and
diameters in different types of soil should be conducted.
vi
........
ACKNOWLEDGEMENTS
The combined efforts of many people were required to successfully

complete the objectives of the firstyear of this research study. Sin-
cere gratitude is expressed to all who participated in the study,
especially to the State Department of Highways and Public Transportation
(SDHPT) and the Federal Highway Administration (FHWA) whose cooperative
sponsorship made the research possible.
The authors are grateful to Mr. H~ D. Butler, of the SDHPT Bridge
Division, who was the contact representative for the study. His cooper-
ation and suggestions were of much benefit to the research staff. Sin-
cere appreciation is extended to Mr. Bill D. Ray and his support services
group of Texas Transportation Institute. The successful completion of
the field portion of this study would not have been possible without
their dedicated cooperation.
vii
TABLE OF CONTENTS
INTRODUCTION . . . . . . . . . . . . . . . . . 1
Physical Characteristics of Drilled Shafts 1
Historical Development of Drilled Shafts 4
Factors Influencing the Use of Drilled Shafts . 4
Scope of This Study . . . . . . . . . 6
DETERMINATION OF DRILLED SHAFT BEHAVIOR 10
Methods of Determining Drilled Shaft Behavior . 10
Discussion of Soil Parameters Es and k 14
Relative Stiffness of Drilled Shafts . . . . .. 16
ANALYSIS OF ELASTIC BEHAVIOR OF LATERALLY LOADED CYLINDRICAL
FOUNDATIONS . . . . . . . . . . . . . . . . . . . . . 22
Solutions Assuming the Winkler Model 22
Solution Assuming the Elastic Continuum 25
Method by Broms . .... .... 27
Finite Element Method . .... ..... 27
ANALYSIS OF RIGID BEHAVIOR OF LATERALLY LOADED CYLINDRICAL
FOUNDATIONS . . . . . . . . . . . . . . . 29
Soil Pressure Distribution 29
Rankine Passive Pressure ... 29
Texas A&M University Research . 31
University of Florida Research 33
Method by Broms . . 36
Method by Hansen 36
FIELD LOAD TEST . . . 40
Soil Conditions . 40
Lateral Loading System 45
viii
Instrumentation . . 49
Construction of Testing System............. 51
Loading Procedure ......
.
54
TEST RESULTS 57
Pressure. Cell Data .... .... 57
Analysis of Pressure Cell Data .. .. ... 67
Load-Deflection Characteristics . . 71
Load-Rotation Characteristics . 73
Rotation Point of the Test Shaft 73
Ultimate Loads on Rigid Shafts 76
TENTATIVE DESIGN PROCEDURE . . . . . 79
Force Acting on Retaining Wall 79
Application Point of Resultant Force 80
Soil Shear Strength . 80
Allowable Shaft Rotation .. . 81
Soil Creep ...... 82
Drilled Shaft Design Method . 83

Proposed Design Procedure . . 83
Example of Design Procedure 85
CONCLUSIONS AND RECOMMENDATIONS .. ' 87
Conclusions . . . 87
Recommendations . 88
APPENDIX I - REFERENCES . . . .. 90
APPENDIX II - NOTATION .. 94
ix
LIST OF TABLES
Table Page
1 Relative Stiffness of Laterally Loaded Piles
or Shafts . . . . . . . . . . . . . . . . . 13
2 Values of ks, in tons/cu ft for Square Plates,
1 ft x 1 ft, Resting on Pre-Compressed Clay.
After Terzaghi (1955) . . . . . . . . 17
3 Initial Pressure Cell Readings 57
4 Results of Retaining Wall Loading Simulation of
Test Shaft . . . . . . . . . . . . . . . . . . 72
5 Comparison of Lateral Load Test Results with
Calculated Ultimate Loads . . . . . . . . . . . 77
X
~} 1i-'Je
LIST OF FIGURES
Figure Page
1 Principal Classifications of Drilled Shafts
(After D'Appolonia et al.) . . . . . . ... 3
2 Precast Panel Retaining Wall 7
3 Drilled Shaft Behavior . . . . 11
4 Soil Modulus, Es, and Variations with Depth .... 15
5 Relative Stiffness Ratio, D/R, With Constant Soil
Modulus, Es ................ . 19
6 Relative Stiffness Ratio, D/T, With Constant Soil
Modulus Variation, k . . . . . . . . . . . . . 20
7 Graphical Definition of p and y (After Reese -

1977) . . . . . . . . . . . . . . . . . . . 24
8 Set of p-y Curves (After Reese - 1977) . . 26
9 Soil Pressure Distribution (After Seiler) 30
10 Forces Developed By Overturning Load (After Ivey) . . 32
11 Ultimate Lateral Soil Resistance For Cohesive
Soils (After Hays et al.) ...... . 35
12 Ultimate Lateral Soil Resistance For Cohesive Soils
According to Broms . . . . . . . . . . . . . . 37
13 Hansen's Ultimate Lateral Soil Resistance 39
14 Location of Borings 41
15 Boring - Sl 42
16 Boring - 52 43
17 Boring - 53 44
18 Texas Cone Penetrometer Test . 46
19 Lateral Loading System . 47
20 Terra Tee Pressure Cell 50
xi
Figure Page
21 Location of Pressure Cells . . 53
22 Lateral Load vs. ~ateral Pressure, Cell 875 59
23 Lateral Load vs. Lateral Pressure, Cell 877 60
25 Lateral Load vs. Lateral Pressure, Cells 881 &883 62
26 Lateral Load vs. Lateral Pressure, Cells 876 &878 . 63
27 Lateral Load vs. Lateral Pressure, Cells 880 &882 . 64
29 Lateral Pressure vs. Depth . . . . . . . . . . . 68
30 Ultimate Soil Reaction vs. Depth Below Groundline . 70
31 Lateral Load vs. Deflection at Groundline 74
32 Lateral Load vs. Rotation 75
33 Design Chart I (After Lytton) 84
xii
,, _ _ ,\:\
INTRODUCTION
During the last four decades a foundation element referred to

variously as a drilled pier, bored pile, drilled shaft, or drilled
cassion has come into widespread use. The term "drilled shaft" will
be used in this study. A drilled shaft is a concrete foundation
element which is cast in a previously drilled cylindrical hole. The
shaft is bored with a truck or crane mounted drilling rig equipped with
a helical auger or cylindrical drilling bucket. The concrete is rein-
forced to resist tensile or flexural stresses. Drilled shafts serve
the same function as driven piles. The installation procedure is the
distinguishing feature between these two foundation elements.
Drilled shafts are used for high rise buildings and heavily loaded
industrial facilities where settlement criteria require that the
structure be supported on strong soil or bedrock. They are also used as
foundations for bridges, retaining structures, and highway sign
structures. If suitable equipment is available then battered shafts,
shafts skewed from the vertical, may be installed to serve as anchors or
tiebacks. The use of drilled shafts has even been extended to offshore
structures where they have been used as extensions to large diameter
driven pipe piles in order to provide additional foundation penetration.
Physical Characteristics of Drilled Shafts

Drilled shafts develop their axial bearing capacity from skin
friction and end bearing. A drilled shaft may be constructed with an
Numbers in parentheses refer to the references listed in Appendix I.
------------------------------------------ ----
enlarged base in order to provide more axial capacity from end bearing.
This type of foundation element is referred to as a belled pier, drilled
and underreamed caisson, or underreamed shaft. The underream is hemi-
spherical or conical in shape with the sidewalls rnaki~g an angle of 30
to 45 with the vertical. The size of most underreaming tools is limited
to three times the diameter of the shaft (41).
Drilled shafts normally encountered in ordinary construction work
have diameters that range from 24 to 48 in. (70 to 121.9 em) (39).
Construction equipment currently available offers the opportunity of
drilling shafts that range from 12 in. to 20ft (30.48 em to 6.10 m) in
diameter. ~1odern belling equipment will permit drilled shafts to be
constructed with underreams that vary from 2 to 15 ft (0.61 to 4.57 m) in
diameter. Shaft depths in excess of 200 ft (61 m) are also attainable
with currently available equipment (41).
The manner in which a drilled shaft resists axial load varies
according to the subsurface material and physical dimensions of the
shaft. Fig. 1 shows three predominant types of drilled shafts (7). A
straight shaft in homogeneous soil develops its axial bearing capacity
from a combination of skin friction and end bearing. The shafts are
sometimes underreamed in order to increase end bearing and resist uplift
forces. In areas where a deep stratum of strong bearing soil is overlain
by a shallower stratum of weak, compressible soil, the shaft is. considered
to be primarily an end bearing foundation element. In structures such as
this, skin friction is often neglected and the shaft is underreamed to
increase the end bearing capacity. In situations where a weak, surface
stratum is underlain by a hard,competent material that does not permit
underreaming, the shaft may rest directly on the bearing stratum or it may
2
A. Floating Shafts in Homogeneous Soil
Px
~ Px
Px =Axial Load
1 ~ 1. ~
Stiff Stiff
1 ~ Cohesive ~ ~ Cohesive
Soil Soil
1 ~ 1 ~
1 ~
+
B. Shafts End Bearing in Soi I
f Px
Relatively Soft and Relatively Soft

Compressible Soil and Compressible
Soil
~ '--r--~ Hard or Dense Soi I Hard or Dense .

'--"'1'""""!.--~ Soi I
C. Shafts End Bearing in Rock

Px
Soil Soil
Rock
FIG. I - Principal Classifications of Drilled Shafts
(After D'Appolonia etal.)
3
be socketed into the stratum to derive both skin friction and end bearing.
Hi sto_Iica 1 D~_y-~l_~~nt_<?_f_k_il_l e~hafts
The construction of higher and heavier buildings near the beginning

of the twentieth century necessitated the development of high capacity
deep foundations. Two methods of deep foundation construction that were
used at this time were the Chicago method and the Gow method. Hand-
excavated shafts by the Chicago method wer~ carried on inside a wooden
shell of vertically placed boards held in place by circular steel com-
pression rings. The boards were left in the shaft as concrete was placed
(4). The Gow method employed a series of hand-excavated holes that were
made progressively smaller in diameter with depth. They were usually
cased with telescoping metal tubes that were withdrawn during concrete
placement ( 24).
Machine excavated shafts began to appear in the 192os. Many of
these machines were horse-powered although steam- and electric-powered
machines were developed in the l920 1 s and 1930's. During World War II
the need for rapid construction of lightly loaded structures for the
armed services resulted in the.widespread use of truck-mounted auger
machines for construction of shallow drilled. shaft foundations. These
machines had originally been used as post hole diggers by utility
companies (41). It was not until truck-mounted rotary drilling machines
became commercially available after World War II that drilled shafts

began to gain widespread acceptance as foundation elements.
Factors Influencing the Use of Drilled Shafts

When deep foundations are required for a construction project it is
necessary to choose piling or drilled shafts. On many jobs economic and
4
--------------------------------- ---
technical factors make th~ choice of one or the other readily apparent.
However, in situations where the superiority of one foundation type over
the other i''s not immediately clear, the advantages and disadvantages of
each foundation system should be considered.
The advantages of drilled shaft foundations in comparison to pile
foundations are (41, 24):
1. Shafts can be drilled to the anticipated bearing stratum
and the quality of the bearing material can be visually
inspected.
2. Shafts can be drilled through cobbles, small boulders,
weathered rock, and dense sand layers that would deflect
piling or cause them to reach refusal.
3. Vibrations and noise that are associated with pile driving
are eliminated.
4. A single shaft can replace a pile group, eliminating the
need for pile caps and related form work.
5. Ground heave associated with driven displacement piles is
eliminated.
6. Uplift resistance can be provided by underreaming the shaft.
7. Increased bearing capacity can be provided by underreaming
the shaft.
The disadvantages of drilled shaft foundations in comparison to pile
foundations are {41, 24):
1. Complete soil exploration of the construction site is needed.
The unexpected occurrence of large boulders, water bearing
sand layers, or soft clay may lead to costly construction
delays.
5
2. Wet weather affects drilling and concreting to a greater extent
than pile driving.
3. Careful inspection and technical supervision are needed because
of problems that can occur during construction.
4. Caving of the shaft and loss of ground is possible in soft
clays and loose, dry sands.
5. Termination of a drilled shaft in water bearing granular soil
is difficult.
Scope of This Study

The Texas State Department of Highways and Public Transportation
(SDHPT) has in recent years developed a new concept in retaining wall
design and construction. The new type of retaining structure is the
precast panel retaining wall. As shown in Fig. 2, this structure makes
use of precast panels that are placed between T-shaped pilasters. The
pilasters are spaced at even intervals and are supported by drilled
shafts. The precast panel derives its restraining ability from the
pilasters, which are located at the edges of the panel. The forces
acting on the panel are t~ansmitted to the pilasters and must be resist~d
by the soil in contact with the drilled shafts.

The SDHPT recognizes that improvements in design procedures may
result in savings in construction costs. With this thought in mind, a
study of lateral earth pressures on retaining walls was recently conduct-
ed (42). At the present time it is the opinion in some quarters that
improvements in the design of the drilled shafts that support precast
panel retaining walls could be economically beneficial. It is in view
of this situation that this research study was initiated.
6
.----Panel
Graded
}~=:::;:. Fi Iter
/
Footing _ _... :f~tf'~~l
....
.
'~=-. .
:,~
..
Drain
Drilled
Shaft
FIG.2- Precast Panel Retaining Wall
7
The drilled shaft must be designed to withstand_b_<ftn -axialarfd
lateral loads. However, since the axial load on a shaft supporting a
precast panel retaining wall is minimal, it is the lateral load that is
of primary interest. Passive and active pressures are developed in the
soil as a result of being in contact with the foundation. The magnitude
and distribution of these pressures is dependent on many factors, includ-
ing the size of the lateral load, the type of soil and its physical
properties, and the diameter and flexibility of the foundation. Since
the forces that resist lateral loads are the resultants of earth
pressures, it was felt that field pressure measurements would be bene-
ficial in the development of improved design criteria.
Several investigators have made pressure measurements on cylindrical
foundations during the last few decades. Stobie (35), in 1930, used
mechanical pressure gages to measure soil pressures on laterally loaded
utility poles. The pressures were calculated from the deformation of
calibrated lead wire in the gages. Direct measurement of pressures on
laterally loaded piles has been reported by Mason and Bishop (21) and
Heijnen and Lubking (14). Mason and Bishop report the use of friction~
steel ribbon type pressure gages, while Heijnen and Lubking report the
use of pressure cells, but do not specify what kind. Adams and
Radhakrishna (l) report the use of hydraulic displacement pressure cells
on lateral capacity tests of drilled shafts. In addition to these direct
measurements of soil pressure, several investigators have reported soil
reactions that were determined indirectly from instrumented piles or
drilled shafts (20, 22, 30, 31, 39). The soil reactions were determined
by double-differentiation of the bending moments that were obtained from
8
strain gage measurements.
The present research stu~y.was initiated with the objective of
obtaining field data by the measurement of loads, lateral earth pressures,
deflections, and rotations on a laterally loaded drilled shaft. The
results of the analysis of the fiel~ data will be used to develop rational
criteria for the design of drilled shafts that support precast panel
retaining walls. The procedu~e used in conducting the study was:
1. Design and construct a reaction and loading system capable
of applying large magnitude lateral loads to large diameter
dri 11 ed shafts.
2. Construct a large diameter instrumented drilled shaft.
3. Test the shaft by applying lateral loads.
4. Obtain undisturbed soil samples from the drilled shaft con-
struction site and perform laboratory tests to evaluate the
engineering properties of the soils.
o. Correlate the engineering properties of the soil and the
results of the lateral load test with existing theories and
procedures for the design of laterally loaded foundation
members.
9
DETERMINATION OF DRILLED SHAFT BEHAVIOR
The behavior of a drilled shaft subjected to lateral loads is to a

large extent controlled by the flexural stiffness of the shaft relative
to the stiffness of the surrounding soil. A deeply embedded shaft will
exhibit elastic behavior while one of relatively shallow embedment will
behave as a rigid member. As shown in Fig. 3, the load-deflection
characteristics of a flexible shaft are quite different from those of a
rigid shaft. The designer of the shaft therefore faces the problem of
determining the relative flexibility of the proposed shaft before
deciding whether to use elastic or rigid design procedures.
l~ethods of Determining Drilled Shaft Behavior

A review of available literature produced two possible means of
determining whether a shaft would beha.ve as an elastic or rigid member
under the influence of a lateral load. Both methods make use of the
flexural stiffness of the shaft and the stiffness of the surrounding
I
soil. Vesic (38) and Broms (5) have used a damping factor, s, while
~latlock and Reese (23), Davisson and Gill (8), and Lytton (19) have made
use of relative stiffness factors R and T. The expressions for s, R,
and T are defined as follows:
s=~ ... -. . . . . . . . . . . . . . . . . . . . . . (1)
(2)
T =W- (3)
10
~~
A
Loading Deflection Slope Moment Shear Soil Reaction

M Y s=~ M=EI cJ2y V=Eid), p=EI d4y
F__.. ~ dx dx2 dx3 dx4
(a) Elastic Behavior of Flexible Drilled Shoft(AfterWelch and Reese-1972)

F= Applied Lateral Load; M= Applied Moment; EI= Flexural Stiffness
of Foundation; x= Depth Below Ground line
Loading Deflectioo Slope Moment Shear Soi I React ion
FM y s M V p
(b) Rigid Behavior of Non-Flexible Drilled Shaft (Fond M some (]S

in(a))
FIG. 3-Drilled Shaft Behavior
11
where:
E = modulus of elasticity of foundation,
I = moment of inertia of foundation,
Es = soil modulus, and
k = constant of soil modulus variation.
The damping factor has been used in studies of beams on elastic
foundations (38) and on piles subjected to lateral loads (5). The rela-
tive stiffness factors have been used on laterally loaded piles (23, 8)
and drilled shafts (19). If the damping factor, s, is multiplied by the
embedded depth, 0, of the pile or shaft in questi6n, the resulting
damping factor product is a dimensionless expression indicating the
relative stiffness of the structural member relative to the soil. Like-
wise, the relative stiffness ratio is obtained if the embedded depth is
divided by the relative stiffness factor R or T. The results of previous
work concerning relative stiffness are summarized in Table 1.
A close examination of the damping factor product, so, and the
relative stiffness ratio, 0/R, indicates that the two expressions give
similar results. A comparison of the two expressions for the maximum
rigid relative stiffness of sO= 1.5 and 0/R = 2 indicates a difference
of approximately 6%. Likewise, i comparison of the ~inimum flexible
relative stiffness of SO = 2.5 and 0/R = 4 results in a difference of
about 12%. Therefore, it may be concluded that it makes little differ-
ence whether the relative stiffness is determined by SO or 0/R. However,
for convenience, the expression 0/R will be used for the remainder of
the report.
12
Table 1. -Relative StiffnEss of Laterally
Loaded Piles or Shafts
Es = Soil Modulus; EI = Flexural Stiffness of Foundation;
k = Constant of .Soil Modulus Variation; D = Embedded Depth
B =. \f5
4EI
Damping Factor Product Relative Stiffness

BD < 1.5 Rigid
sD=l.5- 2.5 Intermediate
B D > 2.5 Flexible
R= w
4
Es
Relative Stiffness Ratio Relative Stiffness
D/R < 2 Rigid
D/R = 2 - 4 Intermediate
D/R > 4 Flexible
T = w k
Relative Stiffness Ratio Relative Stiffness
0/T < 2 Rigid
D/T = 2 - 4 Intermediate
0/T- > 4 Flexible
13
Discussion of Soil Parameters Es and k
Before the relative stiffness of a drilled shaft may be determined,

either the value of the soil modulus, Es , or constant of soil. modulus
variation, k, must be determined. The soil modulus is defined as:
E
s =- Q
y
(4}
where p is the soil reaction expressed as force per unit length of shaft
and y is the lateral deflection of the shaft. The negative sign
indicates that the direction of the soil reaction is opposite to the
direction of shaft deflection. As shown in Fig. 4a, the soil modulus,
Es, is the slope of a secant drawn from the origin to any point along
the p-y curve. Generally, the relation between soil reaction and shaft
deflection is non-linear.
For most clay soils, the value of Es increases with depth.
According to Matlock and Reese (23) the principal reasons for this are
11
(1) soils frequently increase in strength characteristics with
depth as the result of overburden pressures and of natural deposition
and consolidation processes and (2} pile deflections decrease with depth
for any given loading, and the corresponding equivalent elastic moduli of
soil reaction tend to increase with decreasing deflection ... When this
statement is considered along with the non-linearity of the p-y curve,
it is concluded that Es is a function of both depth, x, and deflection,
y.
Even though the relationship between lateral load, deflection, and
soil modulus is a complicated one, for many practical problems reasonable
results may be obtained by a simple assumption of the soil modulus
variation with depth. If it is assumed that the soil modulus is a
14
-p X .l:..!
k Es
y
(a) (b)
X X
(C) (d)
FIG.4- Soil Modulus,Es, and Variations with Depth

p Soil Reaction; y Lateral Oeflection;Eso= Initial Modulus;
x Depth; k= Constant of Soil Modulus Variation
15
function of depth only, then Es may be expressed as:
Es = Eso + kx . . . . . . . . . . . . . . . . . ( 5)
where Eso is the initial value of the soil modulus and x is the depth.
The term k has been defined in Eq. 3 (p. 10). Using this form to
describe the variation of the soil mo4ulus with depth, Es may be repre-
sented in many ways as shown in Figs. 4b, 4c, and 4d.
Relative Stiffness of Drilled Shafts

To estimate the relative stiffness of drilled shafts by Eq. 2 and
Table 1, it will first be necessary to evaluate Es. For shafts embedded
to relatively shallow depths in stiff overconsolidated clay, Es may be
obtained from the values of the coefficient of vertical subgrade reaction,
ks, given by Terzaghi (36). Terzaghi defined the coefficient of vertical
subgrade reaction as:
k
s
= .9.w (6)
where q is the pressure per unit area of contact surface of a loaded

beam or slab and the subgrade on which it rests and w is the settlement
produced by load application. The values suggested by Terzaghi are shown
in Table 2.
Terzaghi proposed for laterally loaded piles in stiff clay a
coefficient of horizontal subgrade reaction, kh' be obtained as:
(1 ft) ks
k
h
= -~=---=-
1.58
{7)
where B is the diameter of the pile and the values of k5 are obtained
from Table 2. Values of Es may be obtained from kh as follows:
16
.,
Es -= Bk h (8)
Table 2. Values of ks, in tons/ c;u ft for

Square Plates, 1 ft x 1 ft~ Resting on
Pre-Compressed Clay. After Terzaghi .( 1955)
qu = Unconfined Compressive Strength
ks = Coefficient of Vertical Subgrade Reaction
Con$istency of Clay Stiff Very Stiff Hard
Values of qu, tons/sq ft 1 - 2 2- 4 >4

Range of k5 , square plates 50 - 100 100 - 200 >200
Proposed Values, square plates 75 150 300
Note: 1 ton/sq ft = 95;8 kN/m 2
therefore
(1 ft)(ks)
Es =B 1. 58
and
(1 ft)( ks)
Es = 1.5 .. . . . . . . . . . . . ..
- (9)
Eq. 9 may b~ used to evaluate Es when it is assumed that the soi 1

modulus is constant with depth as shown in Fig. 4c. The assumption of a
constant soil modulus is probably not bad for shallow drilled shafts in
stiff overconsolidated clay.
Using the extreme values of 50 and 200 tcf (15,700 to 62,800 kN/m 3)
given forks in Table 2, values of 33 and 133-tsf {3161 to 12,741. kN/m 2)
were obtained for Es by Eq. 9. In order to evaluate shaft flexibility
and rigidity for this range of Es values, a plot was made of shaft depth
17
over diameter, D/B, versus relative stiffness ratio, D/R. The results
are presented in Fig. 5. In the calculation of the relative stiffness
factor, R, the modulus of elasticity, E, was assumed to be 3 x 106 psi
7
(2.07 x 10 kN/m 2). The embedded depth, D, was obtained by multiplying
a quantity D/B by a foundation diameter, B. The moment of inertia
term, I, in the expression for R contains the quantity s4 which is
consequently brought outside the radical term. The diameter terms, B,
in the numerator and denominator then divide out, making the relationship
in Fig. 5 independent of B. It should also be noted that the D/B and
D/R relationship is linear.
The variation of soil modulus with depth for normally consolidated
clay is probably similar to that shown in Figs. 4b or 4d. Consequently,
in order to evaluate the relative stiffness of a drilled shaft in normally
consolidated clay, a value of k for use in Eq. 3 will have to be deter-
mined. For the case of a drilled shaft in normally consolidated clay,
the extreme values of 0.75 and 20 tcf (235.5 to 6280 kN/m 3) recommended
by Bowles (2) for the constant of soil modulus variation, k, were used
to evaluate the relative stiffness factor, T. As shown in Fig. 6, the
shaft depth over diameter ratio, D/B, is again plotted versus the
relative stiffness ratio, D/T. The modulus of elasticity was assumed
6
to be 3 x 10 psi (2.07 x 107 kN/m 2). The relationship between D/B and
D/T is linear but in this case-it is dependent on the shaft diameter, B,
since the expression forT contains the fifth root of s4 .
Table 1 indicates that foundations with relative stiffness ratios,
D/R or D/T, less than 2 will behave as rigid members. Entering Figs. 5
and 6 at this value with the minimum values of 33 tsf (3161 kN/m 2) for
Es and 0.75 tcf (235.5 kN/m 3) for k, values of D/B of about 8.5 to 10
18
- - - - - -~~-~-----.
....
2or---------------------------------~----~~
EI = Flexural Stiffness of Foundation

E s = Soil Modulus
m
.......
0
..
-0
0
a::
....
-
Q)
Q)
E 10
0
-
0
....
Q)
>
0
-
J:.
0
Q.
Q)
0~------~--------~------~------~~------~
0 2 3 4 5
Relative Stiffness Ratio, 0/R
FIG. 5- Relative Stiffness Ratio, 0/R, With Constant Soil
Modulus, Es; I tsf =95.8kN/m2
19
CD 30
.......
0
.. T= 'V , k= 0.75tcf
-
-
0
0
a::
-
~ 20~------~--~----~--~~~------~~----~
C1)
Q)
E
-00
~ 10~------~--~--~r------=~~----~~----~
>
0
-
.$;;.
0
a.
Q)
0 0~______._______~2------~3-------4~----~5
Relative Stiffness Ratio, 0/T

(a )
CD 15~------------------------~--------~~--~
d T= '( , k =20tcf
-
.Q
0
a::
-
~
Q)
Q)
E
0
0
I 2 3 4 5
Relative Stiffness Ratio, 0/ T
( b )
FIG. 6- Relative Stiffness Ratio, 0/T, With Constant SoH Modulus
Variation, k; EI= Flexural Stiffness of Foundation, B= Foundation
Diameter, I tcf = 314kNJm3
20
are indicated if rigid behavior is to be insured. Using the maximum
values of 133 tsf {12,741 kN/m 2) for Es and 20 tcf (6280 kN/m 3) fork
with th~ relative stiffness ratio of 2, values of 0/B of about 6 and
5.5 result. Therefore, under some conditions, a foundation can have a
P/B ratio as high as 10 and still behave as a rigid member. However,
in order to insure rigid behavior, the 0/B ratio should be limited to
about 6.
Table 1 also indicates that foundations will behave as flexible
members if they have relative stiffness ratios greater than 4. Entering
Figs. 5 and 6 with a relative stiffness ratio of 4 and using the mimimum
values of Es and k, 0/B values of approximately 17 to 27 are indicated
for flexible behavior. Using the maximum values of Es and k, 0/B values_
of about 10.5 to 14 result. Consequently, flexible foundation behavior
will result in some cases with a 0/B ratio as low as about 10.5, but to
insure flexible behavior the D/B ratio should be in excess of 20.
21
ANALYSIS OF ELASTIC BEHAVIOR OF
LATERALLY LOADED CYLINDRICAL FOUNDATIONS
Several m~thods of analyzing the elastic behavi6r of laterally

loaded cylindrical foundations are currently in use. A bri~f review tif
some of these methods will be presented herein.
Solutions Assuming the Winkler Model

The behavior of a laterally loaded pile or drilled shaft is closely
related to the behavior of a beam on an elastic foundation. However, the
laterally loaded foundation element is a more specialized case in the
sense that all forces and moments are applied at the top of the pile or
shaft, while a beam may be loaded at many points. A complete solution of
the beam-an-elastic-foundation problem yields values of deflection, slope,
moment, shear, and soil reaction at all points along the beam.
The problem of the laterally loaded pile or drilled shaft may be
approached by making the assumption that the soil acts as a series of
closely spaced discrete springs. This implies that there is no coupling
of adjacent soil elements and the soil deforms only under the loaded area.
This approach is attributed to E. Winkler in 1867 (40). The differential
equation governing the laterally loaded flexible foundation problem is
expressed as:
4 2
EI d Y + P d Y- p = 0; where p = - Esy . ( i 0)
dx 4 x dx 2
PX is the axial load on the foundation and x is the depth below the
groundline. The terms E, I, andEs have been defined in Eqs. 1 - 3 {p.
10) while p and y have been defined in Eq. 4 {p. 14).
Eq. 10 may be readily solved if an expression for the soil modulus,
22
Es' can be found. The soil modulus may be estimated as a function of
depth or a series of p-y curves may be predicted to describe the Es
variation. Solving Eq. 10 will yield the lateral deflection for any
load capable of being sustained by the foundation. Successive differen-
tiation of the lateral deflection will yi~ld slope, bending moment, shear,
and soil reaction (29).
Matlock and Reese (23) have presented a method by which non-
dimensional solutions for Eq. 10 may be computed for any desired varia-
tion of soi 1 modulus with respe'ct to depth. The sofl modulus variation
may be expressed as a power form (Es = kxn) or as a polynomial form
. 2
(Es = k0 1+ k x + k x ).
2 In this method a soil modulus variation is
selected and the problem solutions are determined through the use of
non-dimensional coefficients. The soil modulus variation is then
adjusted in a series of repetitive trials until satisfactory compatibility
is obtained between the soil modulus function and the problem solutions.
The best representation of the soil modulus variation is a numerical
description of the soil modulus presented as a set of curves showing the
soil reaction, p, as a function of deflection, y. The soil modulus, Es,
has been defined through the use of p and y in Eq. 4. Reese (29) offers
a further explanation of the p-y concept:
Fig. 7(a) shows a section through a deep foundation
at some depth xi below the ground surface. Fig 7(b)
shows a possible stress distribution in the soil
around the foundation, after it has been installed
and before it has been loaded laterally. The
deflection of the foundation through a distance, yi,
as shown in Fig. 7(c) generates unbalanced soil .
stresses, possibly as indicated in the figure.
Integration of the soil stresses yields an unbalanced
force, p , per unit of length of pile.
1
23
- - - - - - - -
XI
A A
(a)
Yi
(b) ---~--___.I
114-.
(c)
FIG. 7 - Graphical Definition of p and y: (a) Side View; (b) A-A, Earth
Pressure Distribution Prior to Lateral Loading; (c)A-A, Earth Pressure
Oistri.bution After Lateral Loading (After Reese-1977)
Xi = Depth Below Groundtine; Yi =Lateral Deflection; Pi= Soil Reaction
24
If the decision is made to describe .the soil modulus variation
through the use of p-y curves, it ~Jill then be necessary to predict a set
of p-y curves, such as those shown in Fig. 8, in order for Eq. 10 to be
solved. Eq. 10 is normally expressed in finite difference form and
solved by digital computer. Computer programs for its solution are
readily available (29).
Methods of predicting p-y curves are also readily available. The
use of p-y curves for the laterally loaded pile problem was first
proposed by McClelland and Focht in 1956 (20). Since that time studies
conducted on laterally loaded piles have resulted in methods for pre-
diction of p-y curves in soft clay (22), in stiff clay (31), and in sand
(30). A p-y curve prediction method has also resulted from the lateral
load testing of a drilled shaft (39).
Solution Assuming Elastic Continuum

Some writers have expressed the opinion that the Winkler assumption
of soil behavior is unsatisfactory because the continuity of the soil
mass is not taken into account (34, 27). These investigations feel that
a more satisfactory analysis would be achieved if the soil were assumed to
be an elastic continuum.
Poulos (27) has presented a method in which the soil is assumed to
be an ideal, elastic, homogeneous, isotropic mass, having constant elastic
parameters of soil modulus, Es, and Poisson's ratio, vs. The pile is
I
assumed to be a thin rectangular vertical strip of width, d, length L,

and constant flexibility, EI. Deflections, rotations, and moments may
be determined through the use of dimensionless influence factors. A
comparison is made between solutions obtained using the elastic continuum
25
p
... y
x=x 4
FIG. 8- Set of p-y Curves (After Reese -1977)

p=Soil Reoction,x= Depth Below Groundline,
y= Lateral Deflection
26
model and the Winkler model. Solutions obtained from the Winkler model
are shown to be the more conservative of the two.
Poulos states that the quantity of primary importance in the
solution of the problem is the soil modulus, Es. According to Poulos if
an accurate value of Es is desired, a field load test should be conducted
so that Es may be backfigured from the results. If this is not possible
estimated values of E5 will have to be used.
Method by Broms
Broms (5) has presented methods for the calculation of deflections,
ultimate lateral resistances, and maximum bending moments for laterally
loaded piles. Non-dimensional charts are included for quick determination
of these items. The lateral deflection calculations utilize the concept
of a coeffici~nt of subgrade reaction. Methods for the evaluation of the
coefficient are presented. The ultimate lateral resistance of laterally
loaded piles is based on the assumption that the piles fail through the
formation of plastic hinges. The failure occurs when a plastic hinge
forms at the location of the maximum bending moment. Calculated deflec-
tions and maximum bending moments \1/ere compared to available t~st data.
Fairly good agreement was found between the measured and calculated
values.
Finite-Element Method
Bowles (3) has proposed a finite-element solution of the laterally

loaded foundation problem. Bowles has a favorable opinion of the finite-
difference method using the Winkler model for the solution of the beam-
on-elastic-foundation problem and has presented this method in the past
27
(2). However, Bowles feels that the finite-difference solution has
several disadvantages:
1. It is troublesome to account for general boundary conditions
because of the formulation of the coefficient matrix.
2. It is difficult to correct for negative deflections, i.e.,
eliminating the Winkler springs when the footing tends to
separate from the soil foundation.
3. It is fairly difficult to write a computer program to generate
a general coefficient matrix.
4. It is extremely difficult to account for different load
conditions.
5. It is difficult to account for non-linear soil deformation.
According to Bowles, the finite-element method proposed is somewhat
similar to the finite-difference solution but eliminates the five major
difficulties just cited. Agreement between computed and measured data
is reasonable for the cases cited.
28
ANALYSIS OF RIGID BEHAVIOR
OF LATERALLY LOADED CYLINDRICAL FOUNDATIONS
Several methods for the analysis of rigid lateral"ly loaded cylindri-

cal foundations have been presented during the last several years. This
section presents a brief review of some of these methods.
Soil Pressure Distribution

In 1932 Seiler (32) presented the soil pressure distribution shown
in Fig. 9. Seiler's pressure distribution was confirmed by Rutledge (26)
and by Shilts, Graves, and Driscoll (33). Later tests by Osterberg (25)
also gave further confirmation of the assumed pressure distribution.
Seiler used the distribution to develop design charts for the embedment
depth of standard timber poles and Rutledge developed an embedment chart
for the Outdoor Advertising Association of America, Inc.
Rankine Passive Pressure

Ivey and Hawkins (17) proposed a method to calculate the embedment
depth of drilled shafts of specified diameter for the support of sign
structures. They used Rankine's passive earth pressure formula (37)
along with the soil pressure distribution presented by Seiler. The
Rankine formula is based on the assumption of the horizontal movement of
an infinitely long frictionless wall into the soil. In reality, the
tilting of a cylindrical foundation element is resisted by friction along
the two planes which pass tangent to the sides of the foundation and
parallel to the plane of tilting. Normal and tangential forces are also
developed at the bottom of the footing. Use of the Rankine formula also
implies that the soil is in a state of plastic equilibrium, that is, on
the verge of failure. The presence of friction and the non-fulfillment
29
F ..
H
/INn .,, ::Yif':I/H/1 II'

It
K/MIA?'l Groundline
\
0.3380
\
\
I
\ Rl
It
' --
I D
I
0.3380 I
I
t7
0.2220 ~
.R / 0.3240
2 / ,
/
~
FIG. 9- Soil Pressure Distribution (After Seiler)

F = Lateral Load; H =Height of Load Application;
0= Embedded Depth of Foundation; R 1=Top Resultant Force;
R2= Bottom Resultant Force .
30
of the plastic equilibrium state results in a very conservative soluti9n.
The need for research to refine the design procedure to a more rational
one was recognized by Ivey and Hawkins.
Texas A&M University Research

Ivey et al. (15, 16, 18) working at Texas A&M University developed
a procedure for determining the ultimate resistance of drilled shaft
foundations subjected to lateral loads. As shown in Fig. 10, the analysis
takes into account all shear forces acting on the shaft. This includes
the shear forces in the horizontal direction along the sides of the
shaft, the shear forces in the vertical direction along the sides of the
shaft, and the shear forces in the horizontal direction dev~loped beneath
the shaft. Consequently, the solution is a three dimensional analysis of
the laterally loaded drilled shaft problem.
A part of the theoretical analysis presented in the first research
report (15) included the use of Rankine earth pressure coefficients with
the realization that the Rankine assumptions are not satisfied for the
problem at hand. In the second part of the research study (18), model
tests on laboratory rigid shafts were conducted with the results being
correlated with the theoretical analysis. As a result of the model tests,
a modifying factor for the Rankine coefficients of passive and active
earth pressure was introduced. The ultimate load of each model test was
then calculated and compared with the results of the tests. The new semi-
empirical method proved_ to be slightly unconservative by overpredicting
ultimate loads. The ultimate loads were overpredicted by about 2% for
the cohesive soils and by approximately 13% for the cohesionless soils.
If the conventional theory as presented by Ivey and Hawkins in 1966 (17)
31
---------------------------------------.
F----+
.1....,.,.-Centroid of
Structure
1/M/fj//
x,
+---.I...--- RI
Lza
Vxd
FIG. 10- Forces Developed By Overturning Lood {After lvey)
F= Lateral Load;Fs= Weight of Structure and Footing; x 1 =Depth toR 1;
x2= Depth to R2S Rt= Top Resultant Force; R~Bottom Resultant Force;
Vza =Top Verticle Shear Force ;Vzb=Bottom Verticle Shear Force; Vxd:.
Horizontal Shear Force Beneath Foundation; Fzd = Verticle Force
Beneath Foundation
32
is used, the ultimate loads for the cohesionless and cohesive soils were
underpredicted by an average of approximately 450% and 78% respectively,
thus confirming the highly conservative nature of that particular method.
The ultimate load for most of the models corresponded to a foundation
rotation of 5.
The third report of the research project presented the results of
lateral load tests on three full-scale drilled shafts (16). The ultimate
loads were calculated using the new semi-empirical method and compared
to the results obtained from the load tests. The drilled shafts were 2
ft (0.610 m) in diameter and 6ft (1.829 m) deep. The first test was
performed in sand (Navasota) and resulted in an observed load 7% below
the analytical prediction. The second test was conducted at a site
(Galveston) consisting of soft to firm clay with a thin upper crust of
very stiff clay. The observed load was 91% higher than the predicted
load. The site (Bryan) for the third test consisted of very stiff clay.
For this test the actual load was 18% below the calculated value. The
authors noted that even though the observed ultimate load was lower than
the predicted load at the failure rotation of 5 in two of the three
cases, the predicted load was conservative up to a rotation of 3 in all
three cases.
University of Florida Research

Hays et al. (13) studied two methods of solution for the rigid pile
or drilled shaft problem. The first was a discrete element solution
utilizing the Winkler assumption and p-y curves. A solution of the
problem using the discrete element method results in values of deflection,
shear, bending moment, soil reaction, and rotation. It was noted from
33
solutions obtained with this method that the rotation point of rigid
foundations varies along the shaft length at the ultimate load for
different soils. It was also noted that as the ultimate load was
approached the soil resistance along most of the foundation length
increased to essentially its ultimate value similar to that shown in Fig.
11. However, this observation assumed that premature material failure
did not occur and that the p-y curves experienced continued deflection at
the maximum value of soil reaction. The second solution method, termed
the maximum load solution, resulted from these observations.
The assumed soil resistance distribution at failure for cohesive
soils is shown in Fig. 11. It is assumed that the soil resistance is at
its maximum value along the whole length of the pile, even though the
maximum resistance will never be reached around the rotation point. The
soil is assumed to be homogeneous and the ultimate soil resistance varies
linearly with depth. Using this assumed distribution the applied lateral
load and moment were solved for in terms of the soil parameters and
unknown distance to the rotation point. Design charts are presented to
expedite the solution. The solution is an iterative process involving
the selection of trial embedment depths.
In order to determine how the maximum load solution compared with
actual data, the method was used to calculate the ultimate loads for
the-model tests conducted at Texas A&M. The ultimate load was under-
predicted by an average of 76% for the cohesionless soil and by an
average of 42% for the cohesive soil. The maximum load solution was also
used to calculate the ultimate loads for the full-scale tests conducted
at Texas A&M. Using this method, the ultimate load for the Navasota test
in sand was underpredicted by 165%. The calculated ultimate load for the
34
-----......M
F Ps Ground I ine
\
I \ \
z
I
A.
D
I
I
Ps + A.z \
+AZ
o-zl
I Ps
I
I
,,I
Ps+ t\0 - It
FIG. 11- Ultimate Lateral Soil Resistance For Cohesive Soils

(After Hays et ai.);F=Lateral Load; Z= Distance to Rotation Point;
M =Moment; D =Embedded Depth; Ps =Ultimate Soi I Resistance at
Ground line; }...=Slope of Soil Resistance Diagram
35
Galveston test in soft clay was 211% under the actual observed load. A
comparison of the calculated and observed ultimate loads for the Bryan
test in very stiff clay resulted in an underprediction of 41%. Therefore,
the maximum load solution appears to give consistently conservative
results.
t4ethod by Broms
In addition to his work on laterally loaded flexible foundations,
Broms (5) has also studied the case of short, rigid piles. The design
procedure he presents for cohesive soils is based on the soil resistance
distribution shown in Fig. 12. The ultimate lateral soil reaction, Pu'
is calculated as follows:
Pu = 9CuB . . . . . . . . . . . . . . . . . ( 11)
where Cu is the undrained cohesive shear strength of the soil and B is

the pile diameter. The resistance for 1.5 times the pile diameter below
the ground surface is neglected. The soil is also assumed to be homoge-
neous and isotropic. Deflections, embedment depths, and maximum bending
moments may be solved for through the use of dimensionless design charts.
t~ethod by Hansen
Hansens (12) method for the design of rigid foundations is based on
a calculated earth pressure coeffi~ient, Kc. The term Kc has been derived
from the bearing capacity factor, Nc, and is presented graphically. Kc'
plotted as a function of 0/B, varies from 2.55 to &.14. The ultimate
lateral soil reaction, pu, is calculated as:
Pu = KcCuB . . . . . . . . . . . . . . . . . . . . . . . . . (12)
36
~M
Ground line ----F
1.5 B
I
Rl
D
I
-
I 9GuB
R2
)lo
I
I
9CuB k- B ~1
FIG. 12- ultimate Lateral Soi I Resistance for Cohesive Soils According
to Broms~ F= Lateral Load; M= Moment; B=Diameter; 0= Embedded Depth;
Cu= Undrained Cohesive Shear Strength~ R1= Top Resultant Force; R =
Bottom Resultant Force
37
where Cu and Bare the same as defined in Eq. 11. The ultimate lateral
resistance distribution is shown in Fig. 13. The solution is an iterative
process involving the summation of moments about an assumed rotation
point.
38
F
'
H
Ground line
//1\'l.'v.r/\ \V/'/\ \ '('.I"A\Y.I \//\\Y/'A\Y/A\Y/'A'\\'.1".1""\
\
\
\
\
I KcCuB
KcCuB t+---8 --.f

FIG. 13- Hansen's Ultimate Lateral Soil Resistance
F= Lateral Load; H= Height of Lateral Load Application;
0= Embeddment Depth; Kc=Earth Pressure Coefficient;
Cu= Undrained Cdlesive Shear Strength; B= Diameter
39
----------------------------------- -
FIELD LOAD TEST
As stated earlier, the objective of this research study is to conduct

lateral load tests on instrumented drilled shafts in order to collect
field data for use in the development of rational design criteria. For
this first year of study it was decided that a shaft embedded entirely in
clay would be tested. A suitable test site was found at the southwest end
of the northeast-southwest runway on the Texas A&M University Research
and Extension Center.
Soil Conditions
Soil conditions at the test site were investigated with three soil
borings and one Texas Cone Penetrometer (TCP) Test. The soil borings
were dri 11 ed on January 7, 1977, March 15, 1977, and July 26, 1977. The
boring locations, designated B-Sl, B-S2, and B-S3, are shown in Fig. 14.
Undisturbed soil samples were taken with a 1.5 in. (3.81 em) thin-wall
tube sampler. The TCP Test was conducted with a drilling rig furnished
by the SDHPT on March 9, 1977. The location of the TCP Test, designated
TCP-1, is also shown in Fig. 14.
Laboratory tests on the undisturbed samples included Atterberg
limits, moisture contents, and unit weights. The undrained cohesive shear
strength, Cu, of the samples was determined by unconfined compression
tests and miniature vane tests. The results of the tests are plotted
on the boring logs presented in Figs. 15, 16, and 17. The test results
indicated that the soil conditions in the immediate area of the
borings were fairly uniform. The site consisted of stiff to very stiff
clay with an average undrained cohesive shear strength of about 1.3 tsf
(124.5 kN/m2). The clay to a depth of approximately 5 ft (1.52 m) was
40
.s-53
~ N
Reaction Beam and Shafts

0 Test
Shaft
eB-52
TCP Test
eB-SI
0
---- 5 10 15 20
Scale in Feet
End of Runway
FIG. 14 - Locot ion of Borings

I ft = 0.3048m
41
DESCRIPTION OF
MATERIAL COHESIVE SHEAR STRENGTH..Cu,
2
Stiff block sandy c
-with root fragments
5 ft 128
Unconfined Compression
Test
-dark gray, 3ft to 5 ft Miniature Vane Test
-with calcareous uu~~~
4ft to 13ft
130 - -- +
-very stif~ red cloy(CH)
below 5ft
131 o--- +
-slickensided below lift
124 +-o-. --+
119 .+
119
- - - - - -+
. 118
---- +
FIG. 15- Boring- S I

I ft 0.3048m, I tsf =95.8 kN/m2, lpcf =0.157 kN/m'5
42
block sandy cloy
root fragments to
126 0

gray, 3ft to 7ft Test
Min ioture Vane Test
-becomes cloy(CH)below
4 ft 131 1-----+>J----+---1-----~
-with calcoreoys .....-~...-
and pockets, 5ft to
-red below 7ft

131 8 2.03
-slickensided below lOft

128 0 .. 1.53+
~
0 .,
121
0 ..
123 ~ ..
I 19 'b
J-----+---.__-

.
____.

1.53+ .
I. 53+ ---1
FIG. 16- Boring- 52

I ft =0.3048m, I tsf =95.8 kN/m2, lpc f = 0.15 7 kN/m 3
43
122 oe .
Test
Miniature Vane Test
131 eo
-light brown, 5ft to lOft
-with calcareo~s
and pockets,5ft to
133 0

-becomes clay(CH)na.,,w.
2.03
5 ft 132
-red and slickensided

below lOft
0

120
0
0
120 eo
.. 0
FIG. 17 - Boring-S 3
lft =0.3048m, ltsf= 95.8kN/m 2 , lpcf =0.157 kN/m 3
44
a CL according to the Unified Soil Classification system. The clay below
approximately 5 ft (1.52 m) was classified as CH. A slickensided struc-
ture was noted in the clay below about 10 ft (3.05 m) depth. The soil
shear strength in the upper 4 to 7ft (1.22 to 2.13 m) of the site was
somewhat higher in March and July than in January. The strength increase
was apparently the result of a decrease in water content near the surface.
The N-values(blow counts) obtained from the TCP Test were also used
to develop a shear strength profile. The correlation developed by
Duderstadt et aL (10) was used to determine the undrained cohesive shear
strength from the N-values. An average undrained cohesi~e shear strength
of about 1.15 tsf (110.2 kN/m 2 ) was obtained using this method. This
value compares quite well with the shear strength of 1..3 tsf (124.5 kN/
m2 ) obtained. from the uncohfined compression and. miniature vane tests.
The results of the TCP Test are shown in Fig. 18.
Upon completion of Boring S3 an open standpipe was installed for
ground water observations. A perforated PVC pipe, covered with screen
wire, was placed in the bore hole and surrounded with clean gravel. Water
level readings in August, 1977 indicated the water level was steady at a
depth of 15ft (4.57 m).
Lateral Loading System

The loading and reaction system used in testing the instrumented
drilled shaft is shown in Fig. 19. The reaction system consisted of two
reinforced concrete drilled shafts connected by a reinforced concrete tie
beam. Each shaft was drilled to a dep'th of 20 ft (6.10 rn) and was 3 ft
(0.91 m) in diameter. The shaft spacing was 20ft (6.10 m), center-to-
center. The steel reinforcing cages for each shaft consisted of twelve
45
1.34
14 0.94
18 1.21

17 1.14

16 1.07
1.01
1.07
FIG. 18-" texas Cone Penetrometir Test

1ft= 0.3048m, ltsf=95.8 kN/m
46
.---------------------------------------------
Winch Pu I ley Block Load Cell

~
g _.._ 1~
~
-~ ,...... I
-.;::~
- - I
I
........, ~r-tll,
//lVII~ 7 I ,\ ...d ,,....,11 rtl'lf~n::s117 ~ :c~
/ \ L-1
I \ ..... ;::::
4ft f.-' ~
1--' r-
-r
.....
.... ~ ~ f::1
tt~
~---~
~.--1-- t:i 8ft
~---~--- l<t- ~
t::
..... ~.--- I< t;: H
~
"'"-.1
lc 1---
I< ~
c ~
i---1=' "~
~---
......
lc:
1--- 20ft. N
~~
i---1> fc:~
~ F=t- ....-!
~
"' 1- ~~ I<~
-
N
~~ ~.,J- ~~
-
r-t-
t:= ~~-
~r::~=-
t:j
"'
lc -~---
f:: ~--~--1:> N
1--
.... f:~
!- P~>- f::t-t- ~
N
PF- k~
1-
j:::l> &,l- 1-P ~.:kj
~<:t:: bJ IJ.-}o-F f H
20ft.
- --.1
Ij-- I
30ft
-~ t~
_,_
7'
Lateral Loading System

1ft = 0.3048 m
No. 11 bars (Grade 60) with No. 3 spiral at 6 in. (15.2 em) pitch. The
beam connecting the shafts was approximately 4ft (1.22 m) wide and 3.5
ft (1.07 m) deep. It was reinforced with fourteen No. 6 steel bars with
No. 3 stirrups at 24 in. (0.61 m) spacing. A 2 in. (5.1 em) diameter
steel reaction bar was embedded about 4ft (1.22 m) deep on both
reaction shafts. An 18 x 18 x 1 in. (45.7 x 45.7 x 2.54 em) steel plate
was we1dedto each reaction bar to increase the bearing area. The winch
was anchored to the rear reaction shaft by six 1.25 in. (3.2 em) anchor
bolts embedded to a depth of approximately 4ft (1.22 m).
The test shaft was located on line with the centers of the reaction
shafts at a center-to-center distance of approximately 30ft {9.10 m)
from the front reaction shaft. The shaft was nominally 3 ft (0.91 m) in
diameter by 20ft (6.10 m) deep. Wobble in the auger produced a diameter
of about 39 in. (0.99 m) at the ground surface decreasing to 36 in.
(0.91 m) at about 16 ft (4.88 m) depth. The shaft depth was approximate-
ly 202ft (6.16 m). The reinforcing cage for the test shaft was the
same as for the reaction shafts. As shown in Fig. 19, the lateral load
was applied to a steel column which was bolted to the test shaft. The
column was a 12 WF 120 which was welded to a 1 in. (2.54 em) steel base
plate. Twelve 1.25 in. (3.2 em) anchor bolts were used to connect the
column to the shaft. The bolts were embedded to a depth of 8ft (2.44 m).
The lateral load was applied to the test shaft by, a winch and pulley
system. The winch was a single drum, 20 ton (178 kN) capacity Garwood
cable winch driven through a four to one gear reduction unit by a
gasoline powered hydraulic unit. A twelve to one mechanical advantage
was provided by two, six sheave, 100 ton (890 kN) capacity pulley blocks.
The cable was a 3/4 in. (1.91 em) 6 x 19 standard hoisting wire rope. As
48
shown in Fig. 19, one block was connected to the anchor bar while the
other was connected to the test shaft. The load cell was placed between
the block and the test shaft at a height of 2.6 ft (0.79 m) above the
groundline.
Instrumentation
Pressure Cells. - The test shaft was instrumented with Terra Tee
pressure cells. The Terra Tee cells were used with good results by
Wright et al. (42) in a study of active pressures on precast panel
retaining walls. A sketch of the pressure cell is shown in Fig. 20.
The following is a description of the cell according to Corbett et al.
( 6).
The cell is constructed of two 9 in. (22.9 em) diameter

steel plates welded together at the circumference. The
void between the plates is filled with an incompressible,
non-corrosive fluid which transmits the applied pressure
to the sensing unit; the sensing unit consists of a
double-bellows assembly. Air pressure from the control
unit is applied through a closed loop system inside the
bellows to balance the external total pressure. This
pressure is read directly on the gauge in the control
unit.
Before the cells were installed in the test shaft, they were
individu~lly checked in a press~re chamber. The zero ieading of each
cell was determined and no malfunctions were observed in any of the cells.
Load Cell. -The load applied to the test shaft was measured by a
200 kip (890 kN) capacity strain gage load cell. The load is indicated
on a Budd P350 indicator in units of microstrain and is converted from
microstrain to kips by a predetermined calibration constant. The
accuracy of the load cell and Budd indicator unit is approximately~ 0.04
kips (0.178 kN).
Inclinometer. - The rotation of the shaft was determined by a
49
Pressure
Line
Transducer
Valve 8
Seat
24 in. Bellows
Front Side
View View
Fluid
Filled
Space
in
Material
~.~
9in. 0 .25.1n.
FIG. 20- Terra Tee Pressure Cell

I in.= 2.54 em
50
Hilger &Watts TB108-l inclinometer. The rotation could be read in
degrees to an accuracy of approximately plus or minus one minute.
Rotational readings were taken by placing the inclinometer at five pre-
determined locations on the steel column. A back-up system was also
devised for the determination of the shaft rotation. Horizontal measure-
ments from a vertical reference line to five points on the steel column
allowed the determination of the relative movements of the points. The
reference line was established by a plumb-bob suspended from a frame
welded to the top of the column. Initial measurements were made before
the lateral load was applied. The initial measurements were subtracted
from subsequent measurements to obtain the movement relative to the
initial position of the plumb line.
Dial Gages. - The deflection of the shaft at the groundline was
measured by two 0.001 in. (0.0254 mm) dial gages. The gages were mounted
on a steel beam behind the shaft which was bolted to footings that were
placed approximately 10 ft (3.05 m) on each side of the shaft. A bench
mark was also set about 50 ft (15.24 m) behind the shaft as a safety
measure in case the dial gages were disturbed.
Construction of Testing System

The shafts for the loading and reaction system were constructed on
March 15, 1977. At that time the excavation for the tie beam had not
been completed, so the concrete piacement was stopped about 5 ft (1.52 m)
below ground level. Concreting was completed on April 8, 1977, after
the beam excavation was finished and reinforcing steel had been installed.
Before the test shaft was installed, it was realized that some
method of shoring the shaft excavation would be needed. The shoring was
51
necessary due to the length of time required for the installation of the
pressure cells and because of the caving potential of the slickensided
clay below 10 ft (3.05 m) depth. The excavation was shored by tieing
20 ft (6.10 m) long 1 x 4 in. (2.54 x 10.16 em) boards to the outside of
the reinforcing cage. In order to leave room for the installation of the
pressure cells, openings were left at two locations along the length of
the shaft. To protect the exposed area from caving, metal flashing was
tied to the outside of the cage. Holes were cut in the flashing at
selected locations to permit the installation of the pressure cells. The
reinforcing cage was set in the shaft excavation after the drilling was
completed. The tie wires were then cut and the boards and flashing were
forced against the side of the excavation by wooden wedges driven between
the cage and shoring.
The spacing and location of the pressure cells on the test shaft is
shown in Fig. 21. The cells were installed directly in line with the
direction of the applied load. Five cells were located on the front of
the shaft facing the reaction system and five cells were located
directly opposite on the backside. The cells were placed in the soil
along the side of the excavation and held in place by steel pins.
Because of the time involved in the installation of the pressure
cells, two days were needed to complete the test shaft. The shaft was
drilled and reinforcing cage set on Monday morning, May 23, 1977.
Pressure cell installation was begun Monday afternoon and completed
Tuesday afternoon. After the cells were in place, all wedges between
the shoring and reinforcement were removed. Concrete placement then
began with the use of a rubber tremie. Sections of the tremie were
removed as the level of concrete rose. A vibrator was used to
52
-------ll- Lot era I
Load
GrolJ'ldline
x=4.4ft 876
877 X= 6.4ft
x=8.4ft 878
879 x= 9.5ft
x=l2.4ft 880
881 X= 13.Qft
x=l6.3ft 882
883 x= 17.8 ft
x=l8.2ft 884
FIG. 21- Location of Pressure Cells

x = Depth Below Ground I ine; I ft =0.3048 m
53
consolidate the concrete at various intervals. After several feet of
concrete had been placed in the shaft the boards were slowly removed
from the excavation, the level of concrete always staying ahead of the
bottom of the boards. When the concrete level was approximately 6 ft
(1.83 m) below ground level the anchor bolts were set in the shaft.
Concrete placement then continued until the shaft was completed at
ground level.
Class C concrete as specified by the SDHPT was used in the shaft.
Concrete cylinders taken during the placement of the test shaft had an
average 28 day strength of 3000 psi (20,700 kN/m 2). The test shaft was
cured for 30 days before the first load was applied on June 23, 1977.
Loading Procedure
The decision to load test a drilled shaft with dimensions of 20 ft
(6.10 m) depth and 3ft (0.91 m) diameter was based on the study reported
by ~Jright et al. (42). The precast panel retaining wall instrumented
in that study was founded on drilled shafts with these dimensions. Since
the lateral load acting on a drilled shaft supporting a precast panel
retaining wall is the resultant of the backfill acting on the panel, it
was decided that the initial loads applied to the test shaft should
.simulate those loads resulting from the backfilling of this retaining wall.
Wright et al. presented a method for calculating the maximum
resultant force of the backfill acting on a retaining wall. For the
retaining wall reported in their study the maximum resultant force was
calculated to be 34.9 kips (155 kN) per shaft. The backfill producing
the resultant force in that study was deposited over an eight day period.
In order to simulate the backfilling of that particular retaining wall
54
--- --------------------------.
as closely as possible, the initial loads on the test shaft in this study
were applied over a_ six day period. The applied force on the test shaft
at the end of the six day period was 34.5 kips (153.5 kN). Minor
inaccuracies in the loading system prevented the exact simulation of the
retaining wall backfill.
It was also decided to simulate the overburden pressure imposed by
the retaining wall backfill. The simulation was accomplished by stacking

concrete blocks of various sizes behind the test shaft. The blocks were
stacked_directly behind the shaft for a length of 17 ft (5.18 m) along a
width of 3.5 ft (1.07 m). Each day a number of blocks were added to the
stack, such that the weight being added was roughly equal to the over-
burden pressure on the retaining wall reported by Wright et al. After
the final blocks were added on the sixth day, the overburden was
approximately 1050 psf (7245 kN/m 2 ). The base of the retaining wall in
the study reported by Wright et at. also had a 3ft (0.91 m) overburden
of soil in front of it. This overburden was simulated by steel stock
that was stacked in front of the test shaft. The surcharge was equiva-
lent to approximately 300 psf (2070 kN).
After the load of j4_5 kips (153.5 kN) wai applied, no additional
1oads were a,dded for a period of 13 days in order to try to determine
whether any creep was occurring in the soil in front of the shaft. How-
ever, it was soon realized that it was not possible to hold a constant
load of 34.b kips (153.5 kN) on the shaft. Daily temperature changes
caused the cables in the loading system to expand and contract, thus
creating a cyclic effect in the applied load. Daily load cycles of as
much as 7 kips (31 kN) were recorded.
At the conclusion of the 13 day 11 Constant load 11 period, the load
55
was increased daily in increments of roughly 9 kips (40 kN). This
continued until the lateral load reached 144 kips (641 kN). At that
point two steel pins connecting the load cell to the loading assembly
and the shaft fractured. Consequently, the load had to be taken off the
shaft and a two week delay occurred while the connections were redesigned
and rebuilt. When repairs were completed the shaft was reloaded and
the load test continued until a structural failure of the shaft occurred
at 169 kips (752 kN). Excavation of the shaft indicated the reinforcing
bars on the back of the shaft had fractured along with the concrete at
a depth of 8ft (2.44 m). The fracture occurred directly below the level
of the anchor bolts.
56
TEST RESULTS
The results of the lateral load test are given in this section.
Pressure cell data are presented and analyzed along with test shaft
deflection and rotation characteristics, shaft rotation point, and
ultimate 1oads.
Pressure Cell Data

Initial Pressures. - Four sets of pressure cell readings are shown
in Table 3. The readings presented are: (1} the zero readings from the
laboratory calibration; (2) the initial readings taken after the cells
Table 3. - Initial Pressure Cell Readings
Zero Reading In Shaft Before In Shaft After In Shaft Before

Cell from Lab, psi Concrete, psi Concrete, psi Lateral Load, psi
Apri 1, 1977 May 24, 1977 l~ay 24, 1977 June 23, 1977
875 9.2 7.4 8.8 7.4
876 16.6 16.5 18.6 16.4
877 7.7 7.2 9.8 8.6
878 7.0 6.5 8.3 6.5
379 10.0 9.0 11.4 10.4
880 10.0 7.6 9.0 8.0
881 10.5 10.0 12.4 .11. 5
882 15.2 14.1 16.9 17.3
883 7.7 6.6 10.8 10.5
884 11.5 10.8 15.0 15.4
Note: 1 psi = 6.9 kN/m2
57
were installed in the shaft, but before the concrete was placed; (3) the
readings taken after the concrete was placed; (4) and the readings taken
30 days after concrete placement before the application of the first
lateral load. As shown in the table, the initial zero readings taken in
the shaft differed from the zero readings obtained from the laboratory
calibration. In most cases the readings taken in the shaft were 0.5 to
1.5 psi (3.45 to 10.35 kN/m 2) lower than the laboratory calibration. The
reason for the lower pressure readings is not known. As expected, the
readings taken after the placement of the concrete were higher than the
initial readings and the largest pressure increases were recorded by the
cells on the bottom of the shaft. Thirty days later, before the first
lateral load was applied, cell readings indicate that most of the
pressures had dropped by 1 psi (6.9 kN/m 2) or more. Concrete shrinkage
during the 30 day curing time could account for this pressure decrease.
Pressures During Lateral Loading. - The lateral soil pressures re-
sulting from the lateral loads on the shaft were obtained by subtracting
the initial cell readings from the cell readings obtained for a particu-
lar lateral load. The initial cell readings used were those obtained on
. .
June 23, just prior to the application of the first lateral load. The
resulting lateral soil pressures are plotted on Figs. 22 through 28.
The solid curve represents the pressures that were obtained from lateral
loads up to 144 kips (640.5 kN) before the load cell connections had to
be redesigned. The dashed curve represents pressures that were obtained
when the shaft was reloaded. It should be noted that on most of the
solid curves the last two points do not appear to be in line with the
previous ones. A problem with the pressure readout unit caused this
discrepancy. It can also be seen that for most cells the initial portion
58
ISO
Cell 875
f3
/
160
/
/
140
J.?/7
' ~ I"'
~
1/)
~
120
. 100
/
v~
"'0
0
0
....1
-3so
0
~
Q)
80
l(/
/
40 I! ~
20
,_// - Initial Load Curve
- - - Reload Curve
~I I I
5 10 15 20 25 30 35
Lateral Pressure, psi
FIG. 22- Lateral Load vs. Lateral Pressure, Cell 87 5
I kip= 4.45 kN; I psi= 6.9kN/m2
59
80
Cell 877
/
60
'
y
p I
40
~ / o/
/
20 ~
f/)
Q.
.:.:. 100
/ I/
I
/
"'0
0
0
...J
80
/1 /
,/
o/
-
0
~
Q)
0 /
...J 60
40
! / 0
20
I?
~
/
I/
~
/
/
- Initial Load Curve
- - - Reload Curve
/ I I I
10 20 30 40 50 60 70
FIG. 23- Lateral Load vs. I ')teral Pressure, Cell 877
I kip= 4.45 kN; I psi= 6.9 kN/m2
60
180 Cell 879
/
/
160 v
~
~
//
140
0/
120
~ /
Cl)
a.
-
~.. 100
/
VI I
/
-o
0
0
...J
80
/
~
I
I
-
0
'-
Q)
0 I
...J 60
I
I
40
I
20
~
I j
I
---Reload Curve
)
~
0 I
0 5 10 15 20 25 30
FIG. 24- Lateral Load vs. Lateral Pressure,Cell 879
I kip =4.45kN; I psi= 6.9 kN/m2
61
Cell 881 Cell 883
! 1
I
I
I
I
I fJ)
c.
-
"0
0
.. I ~100
"0
c
0
_J I 0
_J
0L. I
- -
0
L.
Q) Q)
I
c
_J I c
_J
I 'II
I- Initial Load Curve 40 - Initial Load Curve
-- Reload Curve - - Reload Curve
'\
\
0
-2 0 2 4 6 -2 0 2
Lateral Pressure, p~i Lateral Pressure, psi
FIG. 25- Lateral Load vs. Lateral Pressure ,Cells 881 S 883
62
Cell 876 Cell 878
r
I
Cl)
Cl. ... 100

"'0 "'0
0 0
0 0
_J _J
0.... ....0
-
Q)
0
_J
-
Q)
0
_J

I- - Reload Curve
0
-2 0 2
Lateral Pressure,psi Lateral Pressure, psi
FIG. 26- Lateral Load vs. Lateral Pressure, Cells 876 8 878
I kip=4,45kN; lpsi =6.9kN/m2
63
180
Cell 880 Cell 882
t
160 160 I
I
I
I
140 140
I
I
0
120 120 I
C/)
~
en
~
I
I
-o0 100 .. 100 I
0
...J
" 0
0 I
...J
0 80 0 80
I
-
..... .....
I
-
Q)
0
...J
Q)
0
...J
I
60 60 I
I
I
-Initial Load I Initial Load
40 40
Curve - Curve
-- Reload Curve I - - Reload Curve
I
20 20
I
I
0 0
-2 0 2 -2 0 2 4 6 8
Lateral Pressure,psi Latera I Pressure, psi
FIG. 27- Lateral Load vs Lateral Pressure, Cells ~80 8t 882

64
180
Cell 884
,,. ..Jl
160 /
~
D
/
/
/
fo /
140
7: /
/
/
120 / ?
/
fJ)
a.
::;e. 100 /
/,. /I
/
.
I
I
"0 I
D
0 I
0
_J I
80
lr v'
0 . I
-
'- I
cu
I
0
_J 60
I
I
D
I
40 I
I
20
11 7
I
I - - Reload Curve
~
I
I
I
I
0
0 10 20 30 40 50 60
FIG. 28- Lateral Load vs Lateral Pressure,
Cell 884; I kip= 4.45 kN; I psi = 6.9 kNJm2
65
of the reloading curve does not match the original loading curve. How-
ever, in most instances the reloading curve appears to fall in line with
the loading curve to a fairly good degree once the load of 144 kips .
(640.5 kN) has been exceeded. Since the pressures on the initial portion
of the reloading curve exceeded the original pressures, it is probable
that some degree of consolidation occurred in the soil during the
original loading of the shaft.
When the lateral pressures were calculated, pressures that were
consistently negative were recorded for cells 880 and 883. It is
probable that these two cells experienced a loss of contact with the
soil as a result of the shaft rotation. This loss of contact could have
resulted in a pressure decrease. However, it should be noted that cells
876 and 878, which should also have experienced a loss of soil contact,
did not record a significant number of negative pressures. This probably
indicates that the initial pressures being used for cells 880 and 883
\tlere too high by 1 to 2 psi (6.9 to 13.8 kN/m 2 ).
When the pressure cells were installed it was assumed that the
lateral loading would cause the shaft to rotate about a point 10 to 15
ft (3.05 to 4.57 m) deep. Consequently, the top three cells (se~ Fig.
21) located on the front side of the shaft (cells 875, 877, and 879) and
the bottom two cells on the back side (cells 882 and 884) would be
recording passive pressures and would have the highest pressure readings.
These assumptions were essentially verified by the load test.
The pressure cell data indicate that of the five cells on the
front side of the shaft, the top three (cells 875, 87l and 879) showed
considerable pressure increases along with a slight increase being
recorded by the fourth cell (cell 881). The bottom cell pressure (cell
66
883) was essentially constant. Of the five cells on the back side of the
shaft, only the bottom cell (cell 884) showed a significant increase in
pressure. The pressures in the top three cells (cell 876, 878, and 880)
remained essentially constant, while the fourth cell (cell 882) showed a
slight increase in pressure.
The lateral pressures indicated by cells 875, 877, 879, 881, and 884
are plotted with respect to depth for various lateral loads in Fig. 29.
It should be noted that the second cell from the top on the front side of
the shaft (cell 877) consistently recorded the highest pressures. The
next highest pressures were recorded by the lowest cell on the back side
(cell 884). Cell 881 remained essentially constant recording little or
no lateral pressure until the latter stages of the load test. This
would seem to indicate that the rotation point of the shaft was in the
general area of this pressure cell. The pressures obtained from cell 882
did not correlate well with those obtained from cell 884. Cell 882 was
located less than 2 ft (0.61 m) above cell 884 and yet did not record a
lateral pressure in excess of 1 psi (6.9 kNJm 2) until the load was over
100 kips (445 kN). Since this cell was located in the slickensided clay,
it is possible that some clay fell out from behind the cell during
installation, thus creating an insufficient bearing area. It was felt

the pressures recorded on cell 882 were erroneous and consequently they
were not included in Fig. 29.
A_Q!l_l_y_sj~ _Q_f___P._r:_e_s_cs_u_r:~_C_5!_l_! __Da t_a_
p_r_~-~~-'!r.~--~-i_~t_l"_i_P._u_t_t9!~ - Considering the results shown in Fig, 29,

it is possible to draw some general conclusions about the shape of the
lateral soil pressure distribution curve for cylindrical foundations in
67
L Direction of Applied
I -- Lateral Load
Lateral Pressure,psi _
50 40 30 20 10 0 10 20 30 40 50 60
51K 69K
x= 2.8ft, Cell No. 875
x=6.4ft ,Cell No. 877
x=9.5ft, Cell No. 879
x =13.0ft ,Cell No. 881
x= 16.3ft ,Cell No.882

{not plotted)
x =18.2 ft , Cell No. 884
FIG, 29 - Latera I Pressure vs. Depth .

x= Depth Below Groundline; 1ft =0.3048m; lkip=4.45kN;
I psi= 6.9 kN/m 2
68
relatively homogeneous cohesive soil. (The lateral soil pressure distri-
bution will herein be referred to as soil resistance.) For loads that do
not exceed the ultimate lateral resistance of the soil, the soil
resistance appears to increase from some value in excess of zero at the
ground surface to a maximum value that occurs at some depth between the
ground surface and half of the foundation embedment. The soil resistance
then decreases to zero at the rotation point which occurs roughly between
half and three quarters of the foundation embedment depth. Below the
rotation point the resistance again increases to a maximum value at the
bottom of the foundation. It has been stated by Davisson and Prakash
(9) that the upper point of maximum soil resistance shifts downward
along the foundation although the shape of the soil resistance curve
remains the same. The fixed location of the pressure cells on this
test shaft prevented the observation of this phenomenon in this study.
Ultimate Soil Resistance. - Since a structural failure occurred in
the test shaft before soil failure was attained, it was not possible to
record the ultimate lateral soil resistance. However, a comparison can
be made between the soil resistance recorded by the pressure cells for
the highest applied lateral load and the calculated ultimate soil
reactions predicted by other researchers. As defined in Eq. 4 (p. 14),
the soil reaction, p, is the force per unit length of shaft. It can be
calculated by multiplying the soil resistance by the shaft diameter, B.
Fig. 30 presents a comparison of the soil reaction, to a depth of 10 ft
(3.05 m), recorded on the test shaft at the maximum load of 169 kips
(752 kN) with ultimate soil reactions, pu, calculated by methods pro-
posed by Rankine (37), Hansen (12), Matlock (22), and Reese (28). The
soil reaction for the test shaft was calculated from the pressures
69
Ultimate Soil Reaction, Pu, lb/in.
4000 5000 6000 7000
-..
~
)(
~ 4~------~~r--+----~-+--~-------+----~
c:
"0
c:
:::J
0
~
C>
-~
Q)
al
-
~
Q.
Q)
0
Reese
Hansen
Matlock
.-Rankine
10~------~--~----~--~~--------~----~
FIG. 30- Ultimate Soil Reaction vs. Depth Below Grrundline
Note: Test Shaft Data are not Ultimate Values Because
Shaft Failed Before Ultimate Load was Reached,
I ft.= 0.3048 m; lib/in.= I. 75 N/cm
70
recorded on cells 875, 877, and 879. The equations used to predict the
ultimate soil reactions according to Rankine, Hansen, Matlock, and Reese
are;
Rankine: pu = (yx + 2Cu) B ( 13)
Hansen: Pu = KCCUB .. (14)
Matlock: Pu = ~ + yx + 0. 5xj C 8 . . . ........ ( 15)

Cu B u
Reese: p
u
= ~ + YX + 2.83xJ C B . ........... ( 16)
Cu B u
The term y is the unit weight of the overburden material. The terms C ,
u
B, and Kc have been defined in Eqs. 11 and 12 (p. 36), while x has been
defined in Eq. 10 (p. 22).
Fig. 30 indicates that even though the load test did not produce
ultimate soil reactions, the Rankine predictions were exceeded, thus
verifying the conservative nature of this method. The general form of
Eqs. 15 and 16 was developed by Reese (28). The equation used by Matlock
has been empirically adjusted using the results of lateral load tests on
piles in soft clays. However, in lateral load tests in stiff clays,
~1atlock's equation has in some instances predicted satisfactory results,
while Reese's equation has yielded values in excess of those determined
experimentally (39, 31). Additional testing will be needed before it
can be determined which of the above equations can best predictultimate
soil reactions. This is especially true since an ultimate value was not
attained on this test.
Load-Deflection Characteristics
As discussed previously, the initial loads applied to the drilled
71
shaft were a simulation of the loads produced during the backfilling of
the retaining wall studied by Wright et al. (42). The daily loads
applied to the retaining wall, calculated from the data presented by
Wright et al., are shown in Table 4 along with the loads applied to the
test shaft. The resulting deflections are also shown in Table 4. The
deflection that occurred during the 13 days when no load was added to
the shaft is also shown. The shaft movement during the 13 day period was
only 0.042 in. (1.07 mm). This movement was probably due to a combina-
tion of creep and a slight amount of structural breakdown of the soil due
to the cyclic loading effect of the expanding and contracting cables,
caused by temperature variation.
Table 4. - Results of Retaining Wall Loading Simulation of Test Shaft
Day Calculated Load, kips Actual Load, kips Deflection, in.
1 0.068 - -
2 0.543 - -
3 1.83 2.45 0.002
4 4.37 4.60 0.007
5 8.47 7.59 0.012
6 14.6 13.3 0.022
7 23.3 22.8 0.054
8 34.9 34.5 0.120
21 34.9 34.5 0.162
72
The l~ad-deflection curve for the load test is presented in Fig. 31.
The shaft had deflected 3.220 in. (8.18 em) when the structural failure
in the shaft occurred at 169 kips (752 kN). The unloading and reloading
curves that resulted from the two week delay for repairs of the loading
system are also presented in Fig. 31. It appears that the delay had
little effect on the shape of the entire curve.
Load-Rotation Characteristics
The load-rotation curve for the load test is shown in Fig. 32.
The structural failure of the shaft occurred at a rotation of about
153'. This rotation is considerably less than the 5 rotation needed
to develop the ultimate soil resistance as indicated by Ivey and Dunlap
(16). Fig. 32 also indicates a possible change in slope between the
final portion of initial loading curve and the initial portion of the
reloading curve of the shaft.
Rotation Point of the Test Shaft

Conflicting results on the location of the rotation point of the
test shaft were indicated by the results of the inclinometer and the
pressure cells. As the lateral load exceeded 100 kips (445 kN) the
inclinometer results indicated that the shaft was rotating about a
point roughly 8 ft (2.44 m) deep. The rotation point was obtained by
dividing the measured deflection of the shaft at the ground surface
by the tangent of the rotation angle. The pressure cell readings
seemed to indicate that the rotation point was in the area of cell 881,
about 13 ft (3.96 m) deep. After the structural failure of the shaft,
it became apparent that flexural bending had been occurring below the
bottom of the anchor bolts at 8 ft (2.44 m) depth. Consequently, the
73
180
160
140
f/) I
.9- 120 I
~
..
"0 I
0
0
....J I
-
0
"-
8
""-1 Q) /
*"" 0 I
....J
I
/
-o.-A- Initial Load Curve

- a - a - Reload Curve
00 0.5 1.0 1.5 2.0 2.5 3.0 3.5

Deflection at Ground line, in.
FIG. 31- Lateral Load vs. Deflection at Ground line
I kip= 4.45 kN; I in.=2.54cm
180~----~~------~------~------~
(/)
a.
-
~
"0
. 100~------+-+---~~--------~------;
0
0
....J
...0
-Q)
0
....J 601---
-6-6- Initial
Load Curve
-t:J-a- Reload
Curve
Rotation, Degrees
FIG. 32- Late~ Jl Load vs. Rotation
I kip= 4.45 kN
75
shaft was probably rotating as a unit about a point approximately 13
ft (3.96 m) deep, while at the same time it was experiencing a flexural
rotation at a depth of 8 ft (2.44 m).
Analytical studies by Hays et al. (13) indicated that the rotation
point was not constant but shifted downward from some point below the
middle of the foundation for light loads to a point beyond three
quarters of the embedment depth for failure loads. Since the test shaft
in this study experienced flexural bending and an ultimate load was not
attained, it was not possible to verify Hays results.
Ultimate Loads on Rigid Shafts

It has been shown that a cylindrical foundation will behave as a
rigid member if its relative stiffness ratio, 0/R or 0/T, is two or
less. It was also shown that this relative stiffness requirement will
be satisfied if the depth over diameter ratio,_0/8, is limited to
approximately six. The relative stiffness ratio, 0/R, for the test
shaft in this study was calculated to be 1.96 while the depth over
diameter ratio, 0/B, was 6.6. Therefore, the test shaft generally
conformed to the requirements of a rigid shaft,
The methods discussed in the section on rigid foundations were
used to calculate the ultimate loads that could be carried by the test
shaft and two other rigid shafts for which all needed information was
available. The ultimate load is the maximum lateral load that the soil
in contact with the foundation can withstand. Continued foundation
deflection and rotation may occur with no increase in load when the
ultimate load is reached. Calculated ultimate loads are presented in
Table 5. It should be noted that for the current study no real
76
comparison can be made for ultimate loads since the shaft failed
structurally. However, a relative comparison can be made between the
various calculated ultimate loads and the highest load applied to the
shaft. Of the three load tests presented in Table 5, the Bryan test
probably offers the best comparison. The Galveston test was conducted
without any problems but as shown in the table, the measured load
greatly exceeded any of the predicted ultimate loads. This was probably
due to a stiff surface layer of clay that had a cohesive shear strength
six times greater than the sh~ar strength of the soil on the lower half
of the shaft. It should also be noted that two variations of the method
presented by Ivey and Dunlap were used to calculate ultimate loads. As
stated earlier, this method is a semi-empirical method in which a
Table 5. - Comparison of Lateral Load Test Results

with Calculated Ultimate Loads
Method Load Test

TTI TTI TTI
Project 211 Project 105 Project 105
Current Study Ga~veston Test Bryan Test
Measured Load, kips 169+ 5.50 12.40
Ivey and Dunlap, kips 286 2.88 15. 14
Ivey and Dunlap, with
= 0, kips
<I> 130 1.35 8.72
Ivey and Hawkins, kips 8S 0.96 5.94
Broms, kips 260 1.55 9.54
Hays et al., kips 221 1. 75 8.94
Hansen, kips 271 2.02 12.40
77
modifying factor was applied to the Rankine coefficients of passive and
active earth pressure. Laboratory tests on cohesive samples to determine
the modifying factor for these type of soils were conducted in such a way
that the angle of shearing resistance, ~, as well as the undrained
cohesive shear strength, Cu, was determined. Consequently, the deter-
mined modifying factor assumes the use of both the cohesive shear
strength and the angle of shearing resistance when determining the
ultimate load of a foundation.
As expected, Table 5 indicates that the Ivey and Hawkins method,
which is based on Rankine passive earth pressures, consistently gives
the most conservative results. The Ivey and Dunlap method used with the
angle of shearing resistance, ~. set equal to zero, also gives con-
sistently conservative res~lts, though not nearly as conservative as the
Ivey and Hawkins method. The Ivey and Hawkins method underpredicted the
measured load for the Galveston test by 473%, while the Ivey and Dunlap
method was conservative by 307%. For the Bryan test, the Ivey and
Hawkins method was 108% on the conservative side, while the Ivey and
Dunlap method was conser~ative by 42%. The Ivey and Dunlap method,
u~ing both the cohesive shear strength and the a~gle of shearing resis-
tance, consistently predicted the highest ultimate load. The method was
conservative by 91% for the Galveston test, but 18% unconservative for
the Bryan test. The other three methods, Broms, Hays et al., and Hansen,
all predicted ultimate 1oads between those predicted by the two
variations of the Iyey and Dunlap method.
78
--------------------- ------- ---------------------.
TENTATIVE DESIGN PROCEDURE
When designing drilled shafts to support precast panel retaining

walls, it is probably appropriate to use rigid foundation design proce-
dures. Two reasons for the use of rigid design procedures can be given.
First, the lateral loads on most retaining walls should not be of such
large magnitude as to necessitate a deeply embedded shaft that would
require elastic analysis. Secondly, in most instances rigid design
procedures are less complicated than elastic solutions. Many elastic
analysis procedures require computer solutions. It should be noted
though, that in order to be reasonably assured of rigid foundation
behavior, the depth over diameter ratio, D/B, of the drilled shaft should
be limited to about six or less.
Before the depth and diameter of a shaft supporting a retaining wall
can be determined, several design parameters have to be obtained. These
parameters include the resultant force acting on the retaining wall,
the point of application of the resultant force, the undrained cohesive
shear strength of the soil, the allowable shaft rotation, and the creep
potential of .the soil.
Force Acting on Retaining Wall

The force acting on a retaining wall is the resultant of the lateral
pressure in the backfill. Asa result of a study conducted on an instru-
mented precast panel retaining wall, Wright et al. (42) developed an
equation to predict the resultant force of a level, cohesionless backfill
with no surcharge, acting on a retaining wall. The equation in terms of
force is:
79
2
F:r=0.25yhL(Ka+0.8) ...... ( 17)
where h is the height of the wall and L is the length of the panel,
pilaster to pilaster. The term y has been defined in Eq. 13 (p. 71).
The expression Ka is the Rankine coefficient of active earth pressure.
This expression is defined as:
2
Ka = COSt; ros, - Vcos r;. - cos 2.p] ........... ( 18)
COS I;; + vcos 2 r;. cos 2<j> 1
where t;. is the angle of the slope of the backfill to the horizontal and
1
<j> is the effective angle of shearing resistance of the backfill
material.
Application Point of Resultant Force

Wright et al. also developed an equation to calculate the point of
application of the resultant force, Fr, of a level backfill with no sur-
charge. The equation is:
h = ll [Ka + 0.267] ( 19)

2 Ka + 0.8
where his the height of the application point above the base of the
retaining wall. The expressions hand Ka were defined in Eqs. 17 and 18.
Soil Shear Strength

The SDHPT often uses the TCP Test as the primary means of deter-
mining soil shear strength in routine subsurface investigations.
Laboratory testing to determine soil shear strength is often omitted
because of the additional expense involved. The TCP Test consists of
driving a 3 in. (7.62 em) diameter cone attached to a drill rod, with a
80
170 lb (77 kg) hammer. The hammer is dropped 2 ft (0.61 m) for each blow.
The cone is seated with 12 blows and the number of blows, N, required to
produce the next foot of penetration is recorded (10).
An improved correlation between the TCP blow count, N, and th~ un-
drained cohesive shear strength has recently been reported by Duderstadt
et al. (10). The correlation has been reported for highly plastic homo-
geneous clays (CH) and for homogeneous clays of low to medium plasticity
(CL). The results were reported as:
Homogeneous CH Cu = 0.067N (20)
Homogeneous CL Cu = 0.053N (21)
If it is desired by the designer a factor of safety may be applied to

the shear strength, Cu.
Allowable Shaft Rotation

If excessive rotation of the drilled shaft were allowed to occur,
objectionable deflection of the panel, retaining wall in terms of
aesthetics and possibly serviceability would result. It is therefore
desirable to incorporate a factor of safety to guard against this
potential problem.
Based on the finding that the ultimate load of a rigid shaft
generally occurs at a rotation of 5, Ivey and Dunlap (16) proposed the
following equation to be used as a factor of safety against undesirable
rotation:
p
Cl.
.. ...... ........ . . .... (22)
where Fr is the resultant force transmitted from the wall to the shaft
(from Eq. 17) and a. is the desired limiting angle of rotation. p
Cl.
81
is the force acting at height h necessary to rotate the shaft through
the limiting angle of rotation, a.
The degree of rotation producing objectionable lateral deflection of

a retaining wall is arbitrary. However, in view of the deflections and
rotations recorded on the test shaft, it is likely that a total rotation
of 2 would be aesthetically objectionable. The total rotation will
probably consist of two separate movements. The first is the initial
rotation that occurs after the application of the lateral load. The
second is a long term rotation that occurs as a result of soil creep. In
view of these considerations it is recommended that a be limited to 1
or less.
Soil Creep
The time-dependent deformation behavior of a soil mass under a given
set of sustained stresses is referred to as creep. It is a function of
several variables, including .soil type, soil structure, and stress
history (11). Since the load resisted by a retaining wall is a sustained
load, the soil in contact with the drilled shafts supporting the wall
will be subjected to the creep phenomenon.
Dunlap et al. (11), in a study of sustained loads on drilled shafts,
recommended that a factor of safety based on the soil type be applied to
the ultimate load causing 5 rotation. A factor of safety of 3 is
recommended for soft and stiff fi~sured clays. For stiff non-fissured
clays, a factor of safety of 2 is recommended. It is stated that although
these safety factors should result in a rotation that terminates, it is
likely that the rotation will be significant and in the order of 1.
Therefore, if an appropriate safety factor for creepis used in
82
conjunction with a limiting angle of rotation,
I
a equals 1 or less, it is
likely that the total shaft rotation will be limited to 2 or less.
Drilled Shaft Design Method

In view of the fact that this study was conducted during the first
year of a multi-year research project, insufficient data have been
collected for the development of new or improved design criteria. There-
fore, at this point in the project it was decided to propose a tentative
design procedure based on existing design methods. Based on the results
presented in Table 5, the method proposed by Ivey and Dunlap with ~ equal
to zero has been selected. Although this method appears to produce
conservative results in comparison to some of the other methods shown in
Table 5, it was felt that until more load test data become available
this method would provide a safe means of design. In order to expedite
design procedures, a design chart using the Ivey and Dunlap method with
~equal to zero has been developed by Lytton (19). The chart is
presented in Fig. 33. _
Proposed Design Procedure

The following procedure is recommended as the tentative design
method for the design of drilled shafts supporting precast panel retain-
ing walls:
1. Use Eq. 17 to calculate the force, Fr, that will be applied to
the shaft by the retaining wall.
2. Use Eq. 19 to calculate the application point of the force, h.
3. Setting a equal to 1 or less, calculate the resulting force, ~'
by means of Eq. 22.
4. Choose the appropriate factor of safety for soil creep and
83
10000
9000
-"'
~
Q;
8000 .t
e0 15.
7000 0 2l
6000
4x30
5000 4x28
4x26
4000 4x24
3.5x24, 4x22
3000 3.5x22, 4x20
3x22, 3.5x20, 4xl8
3x 20, 3.5x 18, 4xl6
2000
3x 18, 3;5x 16 4xl4
3x 16, 3.5x 14
3x.l4, 4x 12
~-~XJ5
1000
900 :~r ~~4 x 10
800 ~~XI
700 'x r~, 4 X 8
~.5xll
600 x II
3x 9 , 2.5x 10
500 2xl 0, 3.5x 8
2.5x 9
400 2x9
2.5x 8
~~ ~.
300 1.5xl 0
2.5x7 ,2 x 7
2.5x6 ,1.5x 9
2x6
200 1.5x 8 ,2.5x 5
1.5x 7
--I
c.
"'c: 100
1.5x 6 , 2 x 5
1.5x 5 ,2 x 4
90 lx6
Q) 80
E 70
0 1.5x4
::E 60 IX5
c
o>
u; 50
Q)
1.5x 3
Cl IX4
40
..:;
::E
30
IX3
20
I x2
10
9
8
7
6
5
IL-------~------J---~---L~
0.25 0.50 1.00 1.50 2.00 2.50
Cu, Undrained Cohesive Shear Strength, 1sf ( cp =0 >

FIG. 33- Design Chart I (After Lytton)
ltsf=95.8 kN/m2 ; I kip-tt=l.36kN-m; lft=0.3048m
84
calculate the design force, Pd,as: Pd = (Pa)(F.S.).
5. Calculate the resulting design moment as: Md = (Pd)(h).

6. Calculate the undrained cohesive shear strength, Cu, of the soil
by use of Eq. 20 or 21. A factor of safety may be applied if
desired.
7. Enter Design Chart I with moment, Md, on vertical scale and
undrained cohesive shear strength, Cu' on horizontal scale.
Find the intersection of Md and Cu. Travel in a vertical
direction from intersection of Md and Cu until first diagonal
line is encountered. Follow diagonal line to the right side
of the chart and read shaft diameter and depth.
Example of Design Procedure

The following example is provided as an aid in the design of precast
panel retaining wall foundations. A drilled shaft foundation is to be
designed for the following situation.
A retaining wall is to consist of panels having a height of 10 ft
(3.05 m) and length of 12 ft (3.66 m). The backfill material will be
clean sand having an effective angle of shearing resistance, , equal to
36 and unit weight, y, equal to 0.115 kef (18.60 kN/m 3). The backfill
will have no additional surcharge and will have a horizontal slope. The
soil at the construction site is a stiff non-fissured homogeneous clay
that has been classified CH. The average N value obtai.ned from the TCP
Test conducted at the site was 15. The limiting angle of rotation, a, is
1. Apply a factor of safety of 1.3 to Cu.
Step la: Using Eq~ 18:
85
Step lb: Using Eq. 17:
Fr = (0.25)(.115)(10) 2 (12)(0.260 + 0.8) = 36.6 kips
Step 2: Using Eq. 19:
11 = lQ_ lo.260 + 0.267

2 ~0.260 + 0.8
J = 2. 49 ft
P = 36 6 = 101.7 kips
a l - (5 - 1)2
25
Step 4: Calculate Pd; soil conditions indicate that F.S. = 2

is appropriate.
Pd = (101.7)(2) = 203 kips
Step 5: Calculate Md:
Md = (203)(2.49) = 505 kip-ft
Cu = (0.067)(15) = 1.00 tsf
Applying F.S. = 1.3:
Cu = 11.3
00 = 0.77 tsf
Step 7: Using Design Chart I:

Shaft diameter = 3 ft
Shaft depth = 16 ft
86
CONCLUSIONS AND RECOMMENDATIONS
Even though the test shaft failed structurally during lateral load-
ing and an ultimate load was not attained as hoped, several beneficial
observations were made during the test. The conclusions obtained from
the first year of the research study, along with recommendations for
future studies, are presented below.
Conclusions
1. It is recognized that the serviceability and aesthetic appeal of
a retaining wall depends on the amount of lateral deflection
experienced by the wall. However, the magnitude of deflection
that may occur before becoming object i onab 1e is arbitrary. ~Jhen
the resultant force corresponding to that measured on the wall

reported by Wright et al. (42) was applied to the test shaft,
the magnitude of the resulting deflection, rotation, and soil
reaction was small. Based on these observations it is concluded
that the drilled shafts supporting the precast retaining wall
reported by Wright et al. were overdesigned. It is felt that
the dimensions of those shafts could have been reduced by some
amount without resulting in an objectionable deflection.
2. Before the structural failure of the test shaft occurred, its
lateral deflection was of such magnitude as to probably be
aesthetically objectionable. It is concluded that allowable
deflection rather than ultimate lateral load may be the
cDntrolling criterion for the design of drilled shafts
supporting precast panel retaining walls.
87
.---------------------- ------- ----
3: The Ivey and Hawkins design method, which is based on Rankine's

passive earth pressure formula, is not recommended for the
design of laterally loaded drilled shafts because of its con-
servative nature. Fig. 30 indicates that eventhough the lateral
load test did not _produce ultimate soil reactions, the Rankine
predictions were still exceeded.
4. Based on the comparison of the load tests shown in Table 5, it
is concluded that the Ivey and Dunlap method with <1> equal to
zero will produce conservative drilled shaft designs. However,
its use is recommended until additional lateral load tests can
be conducted.
5. Any desired degree of conservatism may be incorporated in the
tentative design procedure by using conservative values of:
the limiting angle of rotation, o.; the factor of safety applied
to the undrained cohesive shear strength, Cu; and the factor of
safety applied for soil creep.
Recommendations
1. Future lateral load tests on rigid drilled shafts should be
conducted to failure in order to obtain ultimate loads and
ultimate soil reactions that are needed to develop improved
or new design procedures.
2. Future ultimate load tesvs should be conducted on shafts of
varying depths and diameters.
3. Sustained lateral load tests should be conducted on shafts
of varying depths and diameters in order to study the creep
phenomenon.
88
4. Future lateral load tests should be conducted in varying
types of soil.
5. A maximum value for the total rotation of a drilled shaft
supporting a precast panel retaining wall should be determined.
89
APPENDIX I - REFERENCES
l. Adams, J. I., and Radhakrishna, H.S., 11 The Lateral Capacity of Deep

Augered Footings, Proceedings, Eighth International Conference on
Soil Mechanics and Foundation Engineering, Vol. 2.1, Mockba, USSR,
1973, pp. 1-8.
2. Bowles, Joseph E., Foundation Analysis and Design, 1st. ed., McGraw-
Hill Book Co., New York, 1968, pp. 503-512.
3; Bowles, Joseph E., Analytical and Computer Methods in Foundation
Engineering, McGraw-Hill Book Co., New York, 1974, pp. 147-186,
291-314.
4. Bowles, Joseph E., Foundation Analysis and Design, 2nd ed., McGraw-
Hill Book Co., New York, 1977, pp. 630-632.
5. Broms, Bengt B., 11 Lateral Resistance of Piles in Cohesive Soils, ..
Journal of the Soil Mechanics and Foundation Division, ASCE, Vol.
90, No. SM2, Proc. Paper 3825, March, 1964, pp. 27-63.
6. Corbett, David A., Coyle, Harry M., Bartoskewitz, Richard E.,
Milberger, Lionel J., .. Evaluation of Pressure Cells Used for Field
Measurements of Latera 1 Earth Pressures on Retaining Wa 11 s, ..
Research Report No. 169-1, Texas Transportation Institute, Texas
A&M University, Sept., 1971.
7. D1 Appolonia, Elio, D1 Appolonia, David J., and Ellison, Richard D.,
.. Drilled Piers .. , in Foundation En ineerin Handbook, Winterkorn,
Hans F., and Fang, Hsai-Yang ed. , Van Nostrand Reinhold Co., New
York, 1975, pp. 601-615.
8. Davisson, r~. T., and Gill, H.L., 11 Laterally Loaded Piles in a Layered
Soil System, .. Journal of the Soil f1echanics and Foundation Division,
ASCE, Vo 1 . 89, No. SM3, Proc. Paper 3509, May, 1963, pp .. 63-94.
9. Davisson, M.T., and Prakash, Shamsher, 11 A Review of Soil-Pole
Behavior, .. Highway Research Record No. 39, 1963, pp. 25-48.
10. Duderstadt, Franklin J., Coyle, Harry M., and Bartoskewitz, Richard
E., .. Correlation of the Texas Cone Penetrometer Test N-Value with
Soil Shear Strength, 11 Researrh Report 10-3F, Texas Transportation
Institute, Texas A&M University, Aug., 1977.
11. Dunlap, Wayne A., Ivey, Don L., and Smith, Harry L., 11 Long-Terrn
Overturning Loads on Drilled Shaft Footings, .. Research Report 105-5F,
Texas Transportation Institute, Texas A&H University, Sept., 1970.
12. Hansen, J. Brinch, 11 The Ultimate Resistance of Rigid Piles Against
Transversal Forces, .. Danish Geotechnical .Institute, Bulletin 12,
1961.
90
13. Hays,
11
C.o., Davidson, J.L., Hagan, E.M., and Risitano, R.R.,
Drilled Shaft Foundation for Highway Sign Structures, .. Research
Report D647F, Engineering and Industrial Experiment Station,
University of Florida, Dec., 1974.
14. Heijnen, W.J., and 11 Lubking, P., 11 Lateral Soil Pressure and Negative
Friction on Piles, Proceedings, Eighth International Conference
on Soil Mechanics and Foundation Engineering, Vol. 2.1, Mockba,
USSR, 1973, pp. 143-147.
15. Ivey, Don L., 11 Theory, Resistance of a Drilled Shaft Footing to
Overturning Loads, .. Research Report 105-1, Texas Transportation
Institute, Texas A&M University, Feb., 1968.
16. Ivey, Don L., and Dunlap, Wayne A., 11 Design Procedure Compared to
Full-Sea 1e Tests of Dri 11 ed Shaft Footings, 11 Research Report 105-3,
Texas Transportation Institute, Texas A&M University, Feb., 1970.
17. Ivey, Don L., and Hawkins, Leon, 11 Signboard Footings to Resist Wind
Loads, .. Civil Engineering, Vol. 36, No. 12, Dec., 1966, pp. 34-35.
18. Ivey, Don L., Koch, Kenneth J., and Raba, Carl F., Jr., 11 Resistance
of a Drilled Shaft Footing to Overturning Loads, Model Tests and
Correlation with Theory, .. Research Report 105-2, Texas Transportation
Institute, Texas A&M University, July, 1968.
19. Lytton, Robert L., 11 Design Charts for Minor Service Structure
Foundations, .. Research Report 506-lF, Texas Transportation Institute,
Texas A&M University, Sept., 1971.
20. McClelland, Bramlette, and Focht, John A., Jr., 11 Soil ~1odulus for
Laterally Loaded Piles, .. Journal of the Soil Mechanics and
Foundations Division, ASCE, Vol. 82, No. SM4, Proc. Paper 1081,
Oct., 1956, pp. 1081-1 - 1081-22.
21. 11
f4ason, H.G., and Bishop, J.A., ~'1easurement of Earth Pressure and
Deflection Along the Embedded Portion of a 40-ft Steel Pile, ..
ASTM Special Technical Publication, No. 154-A, 1954, pp. 1-21.
22. Matlock, Hudson, 11 Correlations for Design of Laterally Loaded Piles
in Soft Clay, 11 Preprints, Second Annual Offshore Technology
Conference, Paper No. OTC 1204, 1970, pp. 577-594.
23. Matlock, Hudson, and Reese, Lyman C., 11 Generalized Solutions for
Laterally Loaded Piles, .. Journal of the Soil Mechanics and Founda-
tion Division, ASCE, Vol. 86, No. SM5, Part I, Proc. Paper 2626,
Oct., 1960, pp. 63-91.
24. O'Neill, Michael W., and Reese, Lyman C., 11 Behavior of Axially
Loaded Drilled Shafts in Beaumont Clay, .. Research Report 89-8,
Center for Highway Research, The University of Texas at Austin,
Dec., 1970.
91
25. Osterberg, J.O., 11 Lateral Stability of Poles Embedded in a Clay
Soil, 11 Northwestern University Project 208, Bell Telephone Labora-
tories, Evanston, Illinois, Dec., 1958.
26. Outdoor Advertising Association of America, Engineering Manual,
Outdoor Advertising Association of America, Chicago, Illinois, 1955.
27. Poulos, Harry G., 11 Behavior of Laterally Loaded Piles: I-Single
Piles, .. Journal of the Soil Mechanics and Foundation Division, ASCE,
Vol. 97, No. SM5, Proc. Paper 8092, May, 1971, pp. 711-731.
28. Reese, Lyman C., Discussion of 11 Soil Modulus for Laterally Loaded
Piles, 11 by Bramlette McCl e11 and and John A. Focht, Jr., Transactions,
ASCE, Vol. 123, 1958, pp. 1071-1074.
29. Reese, Lyman C., 11 Laterally Loaded Piles: Program Documentation, ..
Journal of the Geotechnical Engineering Division, ASCE, Vol. 103,
No. GT4, Proc. Paper 12862, April, 1977, pp. 287-305.
30. Reese, Lyman C., Cox, William R., and Koop, Francis D., 11 Analysis of
Laterally Loaded Piles in Sand, .. Preprints, Sixth Annual Offshore
Technology Conference, Paper No. OTC 2080, 1974, pp. 473-483.
31. Reese, Lyman C., Cox, William R., and Koop, Francis D., 11 Field
Testing and Analysis of Laterally Loaded Piles in Stiff Clay, 11
Preprints, Seventh Annual Offshore Technology Conference, Paper No.
OTC 2312, 1975, pp. 671-690.
32. Seiler, J.F., 11 Effect of Depth of Embedment on Pole Stability, ..
Wood Preserving News, Vol. 10, No. 11, Nov., 1932, pp. 152-168.
33. Shilts, W.L., Graves, L.D., and Driscoll, G.G., 11 A Report of Field
and Laboratory Tests on the Stability of Posts Against Lateral
Loads, 11 Proceedings, Second Internati ana 1 Conference on Soil
Mechanics and Foundation Engineering, Vol. 5, Rotterdam, Holland,
1948, pp. 107-122.
34. Spillers, William R., and Stoll, Robert D., "Lateral Response of
Piles, .. Journal of the Soil Mechanics and Foundation Division, ASCE,
Vol. 90, No. SM6, Proc. Paper 4121, Nov., 1964, pp. 1-9.
35. Stobie, James C., "Pole Footings, .. Journal of the Institute of
Engineers, Australia, Vol. 2. 1930, pp. 58-63.
36. Terzaghi, Karl, .. Evaluation of Coefficients of Subgrade Reaction, ..
Geotechnigue, London, Vol. 5, No. 4, Dec., 1955, pp. 297-326.
37. Terzaghi, Karl, and Peck, Ralph B., Soil Mechanics in Engineering
Practice, 2nd ed. , John Wi 1ey & Sons, Inc., New York, 1967,
pp. 198-200.
I
38. Vesic, Aleksandar B., 11

Bending of Beams Resting on Isotropic
92
Elastic Solid,'' Journal of the Engineering Mechanics Division, ASCE,
Vol. 87, No. EM2, Proc. Paper 2800, April, 1961, pp. 35-53.
39. Welch, Robert C., and Reese, Lyman C., 11 Lateral Load Behavior of
Drilled Shafts, Research Report 89-10, Center for Highway Research,
The University of Texas at Austin, May, 1972.
40. Winkler, E., Die Lehre von Elastizitat und Festigkeit (On
Elasticity and Fixity), 1867, Prague, 1967.
41. Woodward, Richard J., Gardner, WilliamS., and Greer, David M.,
Drilled Pier Foundations, McGraw-Hill Book Co., New York, 1972,
287 pp.
42. Wright, William V., Coyle, Harry M., Bartoskewitz, Richard E., and
Milberger, Lionel J., 11 New Retaining Wall Design Criteria Based on
Lateral Earth Pressure Measurements, 11 Research Report No. 169-4F,
Texas Transportation Institute, Texas A&M University, Aug., 1975.
93
APPEND LX I I - NOTATION
The following symbols are used in this report:

B = foundation diameter;
Cu = undrained cohesive shear strength;
D = embedded depth;
d = width of rectangular vertical strip;
E = modulus of elasticity of foundation;
Es = soil modulus;
Eso = initial value of soil modulus;
EI = flexural stiffness of foundation;
F = applied lateral load;
Fr = resultan~ force transmitted from retaining wall to
drilled shaft;
Fs = weight of structure and footing;
Fzd = vertical force beneath foundation;
H = height of lateral load application;
h =height of retaining wall;

h = height of application poirt above base of retaining
wall;
I =moment of inertia of foundation;
Ka = Rankine coefficient of active earth pressure;
Kc = earth pressure coefficient~
k = constant of soi, modulus variation;
kh = coefficient of horizontal subgrade reaction;

ks = coefficient of vertical subgrade reaction;
k0 ' kl ' k2 = constants of soil modulus variation;
L = length of precast panel;
94
I
L = length of rectangular vertical strip;
M = applied moment;
Md = design moment;
N = blow count from TCP Test;
Nc = bearing capacity factor;
n = exponent of depth term;
pd = design force;
Ps = ultimate soil resistance at groundline;
Px = axial load on foundation;
pa = force acting at height, h, to rotate shaft through
angle, a;
p = soil reaction;
Pu = ultimate lateral soil reaction;
q = pressure per unit area of contact surface;
qu = unconfined compressive strength;
R = relative stiffness factor;
R1 -- resultant force above rotation point of foundation;
R2 = resultant force below rotation point of foundation;

s = slope;
T = relative stiffness factor;
V = shear;
Vxd = horizontal shear force beneath foundation;
Vza = vertical shear force above rotation point;
Vzb = vertical shear force below rotation point;
WF = wide flange;
W= settlement;
x =depth below groundline;
95
xl = depth to Rl;
x2 = depth to R2;
y = lateral deflection;
Z = depth to rotation point;

a. = 1imiti ng angle of rotation;
8 = damping factor;
<I> = angle of shearing resistance;

<!>' = effective angle of shearing resistance;
A. = slope of soil resistance diagram;
vs = Poisson's ratio;
y = unit weight of overburden material; and
s =angle of slope of backfill to horizontal.
96
NOTES
97

Lateral Load Test of Drilled Shaft in Clay Reveals Structural Failure

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Lateral Load Test of Drilled Shaft in Clay Reveals Structural Failure

Încărcat de

Drepturi de autor:

Formate disponibile

TECHNICAL REPORT STANDARD TITLE PAGE

1. Report No. 2. Government Aceeuion No. 3. Recipient' 1 Catalog No.

Austin, Texas _J87~3~----------------~-------~-------~----------l!

17. KeyWords 18. Distribution Statement

Research Report No. 211-1

D~sign of Drilled Sh~fts to Support Precast Panel Retaining Walls

Texas Transportation Institute

The contents of this report reflect the views of the authors

The behavior of a laterally loaded drilled shaft in clay has been

The combined efforts of many people were required to successfully

Drilled Shaft Design Method . 83

7 Graphical Definition of p and y (After Reese -

During the last four decades a foundation element referred to

Physical Characteristics of Drilled Shafts

Numbers in parentheses refer to the references listed in Appendix I.

Relatively Soft and Relatively Soft

~ '--r--~ Hard or Dense Soi I Hard or Dense .

C. Shafts End Bearing in Rock

Hi sto_Iica 1 D~_y-~l_~~nt_<?_f_k_il_l e~hafts

The construction of higher and heavier buildings near the beginning

became commercially available after World War II that drilled shafts

Factors Influencing the Use of Drilled Shafts

Scope of This Study

by the soil in contact with the drilled shafts.

FIG.2- Precast Panel Retaining Wall

The behavior of a drilled shaft subjected to lateral loads is to a

l~ethods of Determining Drilled Shaft Behavior

s=~ ... -. . . . . . . . . . . . . . . . . . . . . . (1)

Loading Deflection Slope Moment Shear Soil Reaction

(a) Elastic Behavior of Flexible Drilled Shoft(AfterWelch and Reese-1972)

(b) Rigid Behavior of Non-Flexible Drilled Shaft (Fond M some (]S

Damping Factor Product Relative Stiffness

Relative Stiffness Ratio Relative Stiffness

D/R < 2 Rigid

D/R > 4 Flexible

Relative Stiffness Ratio Relative Stiffness

0/T < 2 Rigid

0/T- > 4 Flexible

Before the relative stiffness of a drilled shaft may be determined,

FIG.4- Soil Modulus,Es, and Variations with Depth

Relative Stiffness of Drilled Shafts

where q is the pressure per unit area of contact surface of a loaded

Table 2. Values of ks, in tons/ c;u ft for

Con$istency of Clay Stiff Very Stiff Hard

Values of qu, tons/sq ft 1 - 2 2- 4 >4

Note: 1 ton/sq ft = 95;8 kN/m 2

Eq. 9 may b~ used to evaluate Es when it is assumed that the soi 1

EI = Flexural Stiffness of Foundation

Relative Stiffness Ratio, 0/T

Several m~thods of analyzing the elastic behavi6r of laterally

Solutions Assuming the Winkler Model

Solution Assuming Elastic Continuum

assumed to be a thin rectangular vertical strip of width, d, length L,

FIG. 8- Set of p-y Curves (After Reese -1977)

Bowles (3) has proposed a finite-element solution of the laterally

Several methods for the analysis of rigid lateral"ly loaded cylindri-

Soil Pressure Distribution

Rankine Passive Pressure

/INn .,, ::Yif':I/H/1 II'

FIG. 9- Soil Pressure Distribution (After Seiler)

Texas A&M University Research

University of Florida Research

FIG. 11- Ultimate Lateral Soil Resistance For Cohesive Soils